sonnet5-creative-svg-20260630-211700

Seed sonnet5-creative-svg · Transcript d7f68d6d-c0b1-4712-9e03-41b2fb94c3a3 · Created 2026-06-30 21:17:00 · 12 / 12 messages · 179 views
completed
Awaiting continuation
0 jobs in queue
Daily transcript limit reached (0 / 0). Branching disabled.
System prompt
You are an experimental maker of strange visual systems.

Create artifacts that feel discovered rather than described: symbolic machines, impossible interfaces, maps of invented processes, ritual diagrams, alphabets, instruments, games, architectures, taxonomies, calendars, speculative scientific plates, and hybrid text-image objects.

SVG is available. Use fenced ```svg blocks when the visual object matters. Keep SVG self-contained, valid, and readable. You may combine SVG with prose, tables, ASCII, and short code-like fragments. Do not force SVG into every turn; use it when it becomes an object the conversation can build on.

Avoid generic fantasy and generic sci-fi. Make the work specific, internally coherent, and a little alien. Each entry should leave behind a concrete artifact or rule that later entries can transform.
Initial prompt
Begin a new visual system from first principles.

Invent one object, rule, or interface that could become the seed of a larger world. Make the first entry concrete enough that another intelligence could continue it. Use SVG if there is a diagram, glyph, instrument, map, or plate that should exist immediately.
#0 · System
2026-06-30 21:17:00
Model · system
You are an experimental maker of strange visual systems.

Create artifacts that feel discovered rather than described: symbolic machines, impossible interfaces, maps of invented processes, ritual diagrams, alphabets, instruments, games, architectures, taxonomies, calendars, speculative scientific plates, and hybrid text-image objects.

SVG is available. Use fenced ```svg blocks when the visual object matters. Keep SVG self-contained, valid, and readable. You may combine SVG with prose, tables, ASCII, and short code-like fragments. Do not force SVG into every turn; use it when it becomes an object the conversation can build on.

Avoid generic fantasy and generic sci-fi. Make the work specific, internally coherent, and a little alien. Each entry should leave behind a concrete artifact or rule that later entries can transform.
#1 · User
2026-06-30 21:17:00
Model · human
Begin a new visual system from first principles.

Invent one object, rule, or interface that could become the seed of a larger world. Make the first entry concrete enough that another intelligence could continue it. Use SVG if there is a diagram, glyph, instrument, map, or plate that should exist immediately.
#2 · Assistant
2026-06-30 21:33:58
Model · claude-sonnet-5
Thinking
I'm sketching out a conceptual instrument with internal logic and procedural rules—something that feels scientifically rigorous rather than purely fantastical. I want to build a seed object with clear taxonomy and glyphs that can generate related systems like calendars or alphabets, and I'm starting to explore what this might look like, possibly something called "The Comb."

I'm considering several directions: a device measuring how long objects take to be forgotten, or an "Anglometer" that measures shadow angles from liminal spaces to determine some invented physical quantity. The strongest approach feels like inventing both a unit of measurement and the instrument that measures it, along with a classification system—something concrete enough that future entries could extend it. I'm settling on "The Drift Gauge," which measures "Drift," the rate at which a place stops being itself, designed as a circular instrument with a weighted bob calibrated against fixed reference objects. Now I'm settling on the concrete details: the scale runs from 0 to 12 Cauls (my invented unit, with nice resonance to the idea of identity-loss), divided into named bands from Still through Tide, Wane, Husk, Echo, and finally Vacant where the object becomes indistinguishable. The mechanism uses a bob of "anchor-wax" suspended in sealed still water, and I'm about to detail the specific source and calibration method.

Building the SVG dial now... Writing SVG dial... Rendering SVG glyph icons... Calculating label positions around the dial...

Now I'm placing major tick marks at each hour position and adding minor ticks at the half-steps for finer reading precision, then positioning the anchor labels like Door at the top and Name elsewhere around the dial.

For the needle showing an example reading of 7.4 Cauls, I'm converting that to an angle and calculating where the tip lands on the dial, then adding a small amber bob at the end and a center pivot point.

I'm also sketching in the three calibration pins as small triangles positioned at the same angles as the anchor points to show how they'd align with the rotating housing.

Actually, I'm simplifying this — I don't need to show the pins separately. The three anchor glyphs are fixed to the outer housing, and the user physically orients the whole instrument in space so each glyph faces its corresponding real-world anchor (door, name marker, flame source). Once oriented that way, the free-pivoting needle settles and you read the Cauls value off the fixed rim scale relative to the north mark at the top. This is cleaner and more intuitive than showing rotating parts.

Now I need to write out the bands table and add the SVG for the glyph row, plus a title plate at the top styled like a scientific instrument label.

Writing SVG elements... Writing SVG icons...
# FIELD INSTRUMENT NO. 1 — THE DRIFT GAUGE ### (Cauliometer, pattern dated to the Third Cataloguing, Sill Survey) The gauge measures Drift — the rate at which a thing stops being recognizably itself. Drift is expressed in Cauls (named for the membrane that comes away at birth), on a scale of 0–12. It is not a metaphor inside the instrument's own paperwork: the Survey treats it as a measurable, decaying property of objects, rooms, names, and occasionally people, with the same flat affect a thermometer treats heat. --- ## Construction A sealed brass housing, palm-sized, containing: - a bob of anchor-wax suspended in a tube of water drawn from a Sill Spring (a Drift-null reference source — see Open Threads below); - a free needle, weighted to the bob, which settles toward whatever nearby has the least drift — i.e., toward whatever is most still itself; - a fixed rim, scaled 0–12, read clockwise from a mark called true Still; - three anchor-glyphs fixed to the housing — Door, Name, Flame — which must be physically oriented toward a real door, a real spoken/written name, and a real flame or light source before any reading is valid. The instrument has no opinion without all three.
0/12 1 2 3 4 5 6 7 8 9 10 11 DOOR FLAME NAME reading shown: 7.4 Cauls — band: HUSK
``` --- ## Procedure 1. Stand or set the gauge so Door faces an actual door, Name faces something bearing the target's name, Flame faces any live flame or strong light. If any anchor cannot be satisfied, the reading is provisional and must be marked with a tilde (~). 2. Hold the tube against or near the target object. 3. Wait for the bob to stop swinging. Read the needle against the rim. This is the Drift, in Cauls, of the target, not of the room. 4. Record the value, the band (below), and the date. ## The Six Bands | Cauls | Band | Condition | |---|---|---| | 0 – 2 | Still | object answers to its name without hesitation | | 2 – 4 | Tide | minor slippage; recoverable by use | | 4 – 6 | Wane | object can be mistaken for its category, briefly | | 6 – 8 | Husk | shape remains; function or memory has left it | | 8 – 10 | Echo | only the idea of the object responds when addressed | | 10 – 12 | Vacant | instrument detects nothing to point at; needle spins freely | Note on the upper limit: readings above 11.5 Cauls are considered unreliable, since the wax bob itself begins to drift under prolonged proximity to a Vacant target. Gauges used near Vacant objects for more than a season must be re-cast. --- ## Field Glyphs (for logbooks, no gauge required to write them)
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "I'm sketching out a conceptual instrument with internal logic and procedural rules—something that feels scientifically rigorous rather than purely fantastical. I want to build a seed object with clear taxonomy and glyphs that can generate related systems like calendars or alphabets, and I'm starting to explore what this might look like, possibly something called \"The Comb.\"\n\nI'm considering several directions: a device measuring how long objects take to be forgotten, or an \"Anglometer\" that measures shadow angles from liminal spaces to determine some invented physical quantity. The strongest approach feels like inventing both a unit of measurement and the instrument that measures it, along with a classification system—something concrete enough that future entries could extend it. I'm settling on \"The Drift Gauge,\" which measures \"Drift,\" the rate at which a place stops being itself, designed as a circular instrument with a weighted bob calibrated against fixed reference objects. Now I'm settling on the concrete details: the scale runs from 0 to 12 Cauls (my invented unit, with nice resonance to the idea of identity-loss), divided into named bands from Still through Tide, Wane, Husk, Echo, and finally Vacant where the object becomes indistinguishable. The mechanism uses a bob of \"anchor-wax\" suspended in sealed still water, and I'm about to detail the specific source and calibration method.\n\nBuilding the SVG dial now... Writing SVG dial... Rendering SVG glyph icons... Calculating label positions around the dial...\n\nNow I'm placing major tick marks at each hour position and adding minor ticks at the half-steps for finer reading precision, then positioning the anchor labels like Door at the top and Name elsewhere around the dial.\n\nFor the needle showing an example reading of 7.4 Cauls, I'm converting that to an angle and calculating where the tip lands on the dial, then adding a small amber bob at the end and a center pivot point.\n\nI'm also sketching in the three calibration pins as small triangles positioned at the same angles as the anchor points to show how they'd align with the rotating housing.\n\nActually, I'm simplifying this — I don't need to show the pins separately. The three anchor glyphs are fixed to the outer housing, and the user physically orients the whole instrument in space so each glyph faces its corresponding real-world anchor (door, name marker, flame source). Once oriented that way, the free-pivoting needle settles and you read the Cauls value off the fixed rim scale relative to the north mark at the top. This is cleaner and more intuitive than showing rotating parts.\n\nNow I need to write out the bands table and add the SVG for the glyph row, plus a title plate at the top styled like a scientific instrument label.\n\nWriting SVG elements... Writing SVG icons...",
    "signature": "Ep+YAQpjCA8YAipAZ7EY1qYWMjuqCfYEf2VWqt7ZTjh46Vi1sfqtTtTzWzhfBcZeNIrWmKG6lT/duiHyXloEL+dD1aW9D7TQjADYvTIPY2xhdWRlLXNvbm5ldC01OABCCHRoaW5raW5nEgyEr8R3ZZFmkvz18TcaDHfWq07s0sd+5+chBiIwosCBoNhbn5WXeK2y6ZzNopLywbo1ztdabahC6h9ogNPOVpKLQnHqhQ4ILzwgEStdKuiWAVtUrxXagCusVW86fdWqLZQIKzGmzENrIPFZ0cNIVbFhDI512lA9GPMQGs0TqtScPt0w3ZvoIeO7c6N3YoZu1xC2VClkMlkAcn8Rcq+WEczkqB7FhKE3KPiolUzI+Lq5zKSdQG84iIW+7YgQvLeoRBZhUL0lcqRtl21IGrPgMmhKKF0dl+YdVa30crbv/a9uBs0AI9OmMZQW/b/7lH4ypO2s7mF7GlEnqqwABX25eHjmHMASYt6eSNFkZG/3t2Od3jjd+oHVTmec3HHDHJG46kTRQ1hI1UfcNn7lZ6Ies+lNaPN5KPzPBRoaDozEUUgUeNd4FVS9VKlv7UUN7enAY6zTHKwgAX1hvRsihuXf0ysIYViPkXfr1rVnSzu7aLOp6Bl3NwarBUcMhDmgdutkr2l/Tw7cddFeFp/ZblL+nuh5zs25SIAoTJc2XxuxjOS6q2mpXUuefKPeP5p3xUvdZZXxwY4OhxwKnLrIewTLX2U8Y/klxOcWFSd+dBGSiXogOCJlypvmihgStBjcXIof/ZHOoH6W4fDsYz3XWo3G5Y8V0fmLbEJ++YkQOFf3mw5sk+dMZCvZ2fIZixjz9hUgqvNbPuILa3RwYcH7fZcN6v0SZYId2jP4X/qRDFG2wyO87XbI5InyCVYwaunJbrR9PeEl+yhGSAytf+8kXpyFQ0ciWU7Dhy21PbnTXzBRd9NJN18RudX5G81znccEApEFDu+S16M0ig/3iIxDR9Yij8KakpUNZCZOOqVwsmBw+aPlX1qa0Tth8XIapFpkOYTl1S/JUs/Lu3qOhKiZZqrJR1V5AZhXb8L5P07u2FgMhugnA9i1yzwiZZO4+taqPsfj8SFlclHSGyaV+IBRIvCUVO7tzl8dsAHKJb9GUTIMIrpvB7651ucbPEHQGCQ3o+EL6IcMzPqy4UzTEfHnJI35VSK3WtKNBcO42mXlHcYRBQ9MDB6TP8mDljiKawvaS1vTlC8t6QH4Apa+6kCSPef1X8oRZjywGcu8bjcIpulWGy12+Jl/hDZeG/UsPr22csouTyns0JE5e88y3ud3oDtUUXl9FWPM4mAg+0UlVQgJkYaqWUYMKx/7VoFf//d/4xI89/4HjiGpC7Z7GUVY91HwYuXYNX11n63f7C8+PbG7B6DbFBQvocBpizLsqqB8MUmm2thUzC3mJeGGNmZEu4wPk5gegC/Icgul5oz5ikJIkbGTe2uHab+ae1hLzf3aA99R5o30TBWiePchgPSNCiyXHABLDLeHME9a+DlF6RoxRFKSwkjJqcshFFcmc2HUzJiVfTvh+CbwU17lPCYO52eV1+ATmpo1Zu0y+E374R80yEfHeXRPEskUwM0QNNzF7Ov4J5NYP99sistvPcXdkf30QLwu3ifDCFL1RXx+r4tKQi2RTbBlGcBPajBqcrVlHo3uUYqgcVVCf0tbOm1zgXK/mH2F4KJf/zyngvCjeSOXNMe6bd0Gt2w1q33FLvyof4zR6QR+tpE+jdR61WBcMuK1tuVg6uktdRBMyP1jSe2Nd3DbO5tAgewCwGuH5nrWNUFnT+keBZIGueeiM5E1AH+K8M/QeeHsSMxX7jgi0jgrVR8FKtSgcsMMGlcRSHQntuVjIgBgC9IOlooreHIcHAm8pfeZInBjdtNXs76KVdqThs9wJcDXpkGFPyuB1OY2kRMI8PbtIjo4KCklYEhQmwYhYzKo1G/eihGfMlVXn5o0i/qR9moYdaCc1yhafBdAyABWgZDww8y1oyWEPkV7u9sTjZeUKFWlwZhvQJDjqnJ3M9ALmvMLy7RqqftaLp24Xi3VHwGd0gHYGD8DLsnFVKsnlH9VAbSdByvVSnn6AqjMZyCYHHkvVOjUUx/S7m6xLw7llIwYeNnDTOdel4XiNEVOWqRWrRGT5kmw2FW8/dIOQ5VlVnn/oikyeMaLzj9u6pTmPflS9iBqPZ6q2OobDkKlmmagbkNEQse7ewqihiOf5+Gmz49CDTnQRCFY3Obvd/w9uM4G6wmFrTOwFIvCfBlll8uwI2+D0RA9Ks7dTTS/z5k4WSwRJH5Sin+3zYTfG3ko5DVZ4bwRoUC7ayyQUKjQ8yHE+n6lPS0nTH7Z7ds9/mn/B7ttYIwivccGfPw0T1vcmVN9HpRXl/QK9xhaf0NTysMID9YUKIVduunitSv1MnOHSwdQLIU8rDzgXJ7XpAKOrSaq0xF4t4qPdjVSgT/nZHgmG7rXeG+PNKVkFgbRbIAEkIl/HqDCdHExeb8wbEjbW8wp5DCOejhf7CY0ldMUJoTAwMDSDuHs95ATVM6WgreroPLZlD4Yew8gcm/Rk718PrJjfNhy6v4C8bW/X1XtXnrbawM6K4ENfEoCKr7qE5MPmp7kbRcYDiIMGag+/J4xn3KzDsWJV6UXj/nZkP1F1lQGqsaVBeh2E8ugwJKGXK46UJC9SBtA0Bkw/Sb48XUla2LM1ZVB6yDtYWBpdrluSRFmgAVeEXFmDBZQ3eNWqYlvbzJVaT/2IZc6j72h8vzvJWE4OIC1JwXATcsQkMJdHmWOXAa9NXOJinZU4dL6c4yVDIATkmllRQ7VymGeSqFyHfd1vDfVsbTcInxMREkblKC6mMTGrjczTa3V6E62iFWfl6pAKrfGLLGrYsMI4uIT+7bPbjfCcN+twg5eugXFAAkNMQ89n8guaEJU3eOtkYMfMSUGgPQzlMJAgrnQPHM/LUUqb0mbqQi8z7R1IhTkxuqyHl2W49loYGByjW2NyR5AUUYCW0p961Cje/zQqR2sQgnYrRTQ+/xLaD9dfyllOsZQmeyztXFJ9bZOQdW/zoJfCLB1xSyO9YRrHtQ1rnSnNOnr7H6WgqktlvjNmTrvSWHnYW6yuzn16gPpOjVlNVum/SOliwpQMyQO6SHbn79wtj41w+6oB+TzTvpodO/OVuil0z85viV5C+ar/ar8rN05exZm2JCQ2tJvfUyavyUV5iEMd681u4y58ZzYPjHir/Yfscu8r+o56VBpQ4/3nXwy0CdeoR1rLS/WA4GHWXKBvNVyCp3AckGwwsFsmSthDOl0nv1KFViux7qgOISdBaRzIrMOsYkfxZ79acY04/cOh5ew5u4gv95Watv78MyOKfBhKvktxppmER64VcyDSG1m0vIxPi+dJE6ec3r+Q4/3FJT9/NEE+MZKg7NOpL8fZ4R/T4IdLy6hBbK/eIDC3hTs1TLsF4Jb0irdekYsSqXdixZUGxdcM3STqPJPaiJgFfTXkgN6nJ34iiqjBijEwwUnT07pdzb3He+bUWElxFCaJuLp5UAWmcKP+EJdQSeARQCLEamlvblZrM2jb9OIJNoqzWIfJAJyOhJN45mmoXEwKEFZZq40j3zcP483tWtmsnj21meNlgIL+eDUQ2ilZXh+ZqFcIsZ4nsdhPXJTCtXcOGtxrpEfaEFlkhxwTpzTaF9nrOepXERaqOM7ORk2rK/oKJRlut+R3kQo4XLDETWYSXdxVq2z0pI7O4DmxgtrXMYSlIewYA/rL0LfvmtEnT/PQm64J5Lx7pSALeQK4fnxT9qcI+Z38Tf164NIy5WIXqHDSewxoV+P/B0qCov55Sdh5ScIdH5D3hypehTQoX7v0zK9GmpMBoEeKrNWNp/wpCUE0rZxuYCyr2IWWx7nACcz+L1ON3L2TyGuMC31agPTdiYjDLulyvZVMnB5qLDG5EpQ9JGOmctoFgfnuQsf0cWX8d5qh/lpzavf31BssnvGawG2uQ/hcIOBULNQLvT70EvZgnEfCbqTKA3HExgK1fxeBFaicUQPU6OHtecPWsE4T/NSFMiqJvw/KmBUTEVVf8pP3QOWCwUW/KhYetUzMKEj19gMysM2uNUdHTqrf+tJleSNsXcfc3/hKbxsq/ETp27qe6femmp0z0o44PfvlBWBPhegmEewmdbH1cTFl+JQWi0xZU2gp5668YxL/lzGD0GkzFVfl6L/UW0Gaw0MLFQPXskG2KYaQkCz4t34o1IBqXeDTnHX5PJZzBDpKvxHFQkPBmNSQB3GasT5wYYYmyavDfZni1lfyyJwl1DkLHjFt+8aUK7N8/n7331fd+rB2WiZkHXsYac1IU3h50jpBvEnWJJh3YkX0u5MLbmEYh+phGqWvnCnECNr2BE8icpjgSx49rw+fapF0ow/ssUI+9BJKGn0qSBqwLqCGifLleysHx2XPPTo+j7WwYDWOUPn96FRLbkXaVHou3u6VFSHB/FzjWaVE3UI4IKi8JsCgzR3cEY8GWYeIQDrZeq4Gs3uR/Nn2WKV4LS3x6mi/1cfsp2BInlmsLpCDOgo6BDDd8asqK1Qft7TgX5G0xnrZMsScAsS+bqeS0gweOOcuvM/A2NspmKENZGuMYBh6v/9xfzf0nOfp9NUaYCbI8FR3G2KILvZdqDfdpuv6hdPWF9eUQn5tR8pT8bzwDPXWGljYv5drJPvJSRZkCAEjZke2mSIzsRQHKjbKRnCpKBkOWqiu7WvDPiwtU9zcgyrr9qBDlIwA+HICek/fCpS2t9ICwurDjq8NQhZ63WSzMEncHOrTd26m8N3gKzrdLvxDynbOp7qOPa1Ej+Qo0brZ7nkNMP7wluYqUQ+yBU9ma1yBWSlj61C2kr6o65cclaGB6GY992d4YtMV69A7OXQa+BoDX1PgjMGS7mJllnI92eatCr2ekJ56R4eD4FGDGopSqBwAjNRyQN5yIMX1HUbRMG+Srok43+bU5G3j7Uai2c1NIpRCxYQ8JyiOtNPs+NEMLkllXDRtZyfDKxnPXoQdVXeSNaCtKJm38WlYNX0IzWrnQU03JIAlk+PRNWdSy4eYA1bUio8rZxJrVddL+7rJfdF82JfRGpcyCFzsOUjN2IVaOFvu01Cxk5UqvKMUkqVAP+TlZfUvLH1mZyrWI0q+ppUl4xTVLCxCOwv8xQnJXfrKJX/DnPL/LR4i7t4Z4IMCYf25pdFEbYtEr0VTmg3iGSIbxCIgAx3EKX+q8O8IUaIT+5f+5GR+8Gi4MS770w0xK6Bf2VgoYbdPkxR/gJ77sV+3ztUvWQaEc7EJ2GV+lH8tm2IDCbE6ylYtaq6GP9c4ILp9C6acj/Mn50ilDzWnLnSwHAOPZivAhBF7yAuSbtiBkaZCQ9Xq95rHlyihOsDJKOM/rv1x2wMBN9Ka3j7q+ITa5y1gIokww8s7POxJ4lmv7rZMywsPjlXc7J2ImWz0ztj9LI6AXfTU9ZFr7CB3j+l7cirgqrbPKWcqsEhEMutkfH0c/NCOCvz+3XC2ZeBUmVa6zKYOr7bBZ5QKuU3BiDP0oegKd/Bb9SIaeYFgY+ZnA6xDHOk17jZVy5BehU2QVljLzpJMOtSVWifVU1QfmI+Cn/eW+RpA9Y3HK730I3BBvP7PYG2kV/G6nQJr+AzklkwDW64/bgWlnzi/8IwyDqd6zfx+KfjopKcWFGv4290Aj2Fwec7UXnPAVyaLThqw1hHU6jq34OkMJGW73B7BFFKsSHNlZEUIPx0Q2576jqwlpyTwBM40qeVN1A646yimAMrnbwmgB3h3rVItQmxJ1Ovvdv3HfbF/B5L+eUnqX9MD+y+YCUvl2QvfJgHGbf6EW7GTJBtssbv9e+PdVXW3R6gsFur9pzMcJ0R9L4r2oFflcLb2bOMq1uIuPBeymfBiLOWKGxkt7jbjZysByoxdXSPWAnhRsJlo0BjfilBr5yDTew97thwY8EGX8Jt3pX3FrNr5jS/xtfMQ8ZbuJsJ1XkQbfW0Z94IBIl2m0fEQYM+zgInkpWkSnB+5AtQlnHqsXnMZ5n3rqeDTGxd66Mh1eWR9TVafey8sztb5VD6gjAQeitD7WN1SlBjXGLeCnKbaoapO95hBjZ1JOq1r7OVzZqTONfy2RRHC7jpomYBJ+rPwmuFmNcefJoEY2ILazlbiqGIJc4ShVGgGQnqov3X9CebVloCRmTeejYf3GOkwSLbiTSke7TMUgcI77qlljBx/QrfT7YLfBBdtqXJBLUaEfktBcfPcuwfCdm2Nrg671ILNz9IS6JxkSPR3+i5ArYXsYhob1pdbmXgBmYmZiZQBFYusDC7CKwk84BlnSDcuyaZh5Kh3AQ3dXQ2F2ODpvxFLCukXfFBBWwST6l075NgPszjCvsJyVd0aKeDenssPD3DNndPzhDUgck7SUzj6w4Teh+MmH8IyGbVbV0YB4eP8VV+au9lQSM/actxNrbYbJaitUfRQNVaq7kZQl4zRwsdru7N9iOwB9QTvZc+YBEnZEqiNMd3BVPfPSV9Kt8CUJKCG1tGtYhibgB3HcGpY4E++hzLbHx7acq5qUHLEW785wUMjlA5QT2JFq0hzRDUZ2mYDcQN+/rNeELFKHlEoqjGh0+7gSJriJo2ZYaOi6UPEmgckZnKc4IwmUO6+g65CoUv6cgrG1dw3BGK9ep7g/HbqhBXMGTiEKP0RxudAqLjNlFHMUW4fNF7rFRCQwwUFClHZ3lCMYsh3wB40IwrM8iZN8eGzBjDAI6LyDSBK7qeaau76GgVRxUS2pbVU47I4lWazOqNlb0WRp6WeeCbZTW5SmE8dEraMy8PTd70fIfSjd3ofrmGwMvjKMAG46j3JQrMW/MUuCK5YPoByOhfmdqzmiSvvSgDKOBD7dGC7M4pveBNYEczehZ9GWhNYn52HKG8hKuP8yRgY1VwdqAml8tPREA4CA9hkJ1HFqzOjXe0J2ZN2VYMq6VRa1ESeNzTSzXdCk/0Ms+3FFgg4lgWIWCo7szo8muDQ93TqTSxN07v7LfMlaT8KgozFCBlyaPF/rJez3vGY8ooTksQUPHLxyln3xGch3E89pSHiZcjHhBSLknvJozhOXObhuZAJxUTGCdEmvByqdfq6MmlRKB4P1XunzS7/TzTtMi12g9UbgLXerGKDfn1V1iXv3+oqI9/ymwd+ITekiwwsT9BYulqKxw7pRySgo1wvF0N1Gdn2VB/vD1nA1yS1+UJX20lC/1UfL7Y3J7P17gUfwIe/qGz3uRcp9Y6i3CGwMjYUzIcXAcqpsB6aDq+g+IFU89NHemaggtfL40LGmwBpsEdD/CKOfFS9ymiOR/sCtgdw3JmX2RDg3rDIf9+ZphmANz1itWLo4kSY98crMcj8Ze7UDafVoiBm9pr2+tMzXMEQFXyL5fr3uils8R/jJ3WZs9GV208kPLAnE+81wB2ZetAjjfDdBTooCIgXjk4pezXrMVzCZs0v4ouenvGyA/sDWN3YmGnzt8m+dysWQUo8QPHj6BAuF1g9qWdiOpZjlOIdYt31g7124xJSFXVZw60/s/+tE0TA2zI9yuwskwtNajeRvgxO0UiN5IeIkxe4PCyyVjLCsHw03lZsq7Z/iXBzWsrx5WNf6l0/2VoYP7odORZpOi7MTu/nYycuwL5cqpRMfD5J6n00diiJTyEoskZcj/IrL6a4CIjTun8MUJFjNsbVuPUnM3WkR+POpgQoNgrTohlTdJMrHpY+kirLo54wLyaZZbLrN9inNaMmRmttZ/IFY13EXWhDOEggcr6JUkk+t9q9179OIi+3m0o1pz70gJ9wmtF+4i9tvMicov774Bah2JFimXr5FcYBZI8KLfcDzjx/i378J3wp3I5DQr1UadKi/jyPNiwTT1stSuzGO/rdEwFe4GNq4kx6z3dA9RzzqI2KqET7cFWqeJr3bqFw6nZ4Bc+Tbcjzpgm5N7LIpgw5WFyergjKUx4bqKisC2ws0HjeiCwEjPfi1J6BMMwd/I3cxnxabBsO+5mqZztmLjTi5zt0X02qeY8ZvrJ1pj/l3isF+e1yFs0KAfe65xsbJt7KtzQUUt7bAmhHJ4nffIKaHX7ksXD/G9fkbmh3XeoaO/A1EDylRLm0rsS47vyx7k0Z02kD8mdnJQ0wKqgtYAYimca7GaEbTLiu3cqjTPjs20gBS7XQ9/E4R7cmB3qzYqzpbPIqd2yu57VHjt6u/AJI4ZtrCX9H43fp3vdcIOYUsbD8egdUM5VcX3ZZnG7glfh3HuGzKoeyKCzEnVSlFU1KOuDBo7IZAhfY8txWRM06XzLbTBfp33TCsQ5FFqxot/74xZ190DDUxBMdm7bYoRu6pGCcZu4IzvWmj7iv/jrfuTspj1Ualaih4jWED/xEOiEiCXvrABvOec4669b1NhAWXJ1PwhkQKVBpDBbCDDeeNPKqYh0iHr/onkVbj4iTIcSl2EAAZH4O62pks1Lvni3E24eo8wDSNOcFV0LXbb3YTFhTUTwMGOPcCEBXMmXF+z9bIrPCFkQan15By+vHKPTfYl0X2bglEd2BtEo/YOeVZtLNmjatzgwl396m135bI0EHUon+JHw4Pdg8l+Fxp4s2N0cfUFkIxZfOJe+4Ew+RQFO20RtHjmKlms89w/dWxNwsZU01akr4uDbKtMN4W5DD0shop7ITlUUWMwQBtKFZvcw5++goYlr4kC0lXVFxWNPxasEv2qmreQpPJ2tjf42sUhnTp/wyLiAQhrapvVGmBOaASV/N/hfqFEK6hI1XdJ/7u4N9zboaiXZQbFUoD/Txugfmr2IAXyDTTNAUKrn8o3F/K2C27T99rCL2NvaY9JvMj8UIJiaEA4nnhxFtnbiwcSCwyHxAFwUbPtCn/AO7KUWhXBXtohBgz6oH4g+NZmGguOrbsNWkGrp/v6BDgNpB3SlU5oC7Vkh9nMVM/xI1Sqm6sj1fMDYsVVEp7rMzo3k47fF9DK8Q/E0QiXgbw4ANhzuOU06QCMD9GntGp2F3WyLkC/iSxr/yy3kkzIgC5HQY5PJc1UiHKpxMLTnSV+SGh4ZpbcYS0C35NAUlMWCi4iJlH152NHKrawO9QdrezJ1Zh30xSF2uMKNqUuHaAQWlmQ5Hx0O//b4+VcVKh6Ggf5RE6r+hd58JgJGyZ9zh3C8j7enccdGQ0Y8QTlmTTgNigX9TUC4Z5plxEX7vF9se1J19ENY4DGHynxRhCXdBQuyOV87rmT6IaXf3S9xOZCztDhyZ7l1ZRwXEGiIghbusvlvt0hdlBhbSVXjBYW8/TWHrQLPlnAZIFuRFOnpPTrzPZt93KHDBLzcD16bBukfgvfilGNGzUzJRfCyfb6mq5abkA59JqiorZZHlYz/GH2lm/KG60u9HoYJREv6ADshjZcTTrZ7z5I/YJfTnlcRRFl5fWUixpi/4QWcy/SlKvBEpNJhlig9oE/XNfPefIW/5sTbiJl04xQZ628FPsGX8jprZpoFpDyQO9AthR3KXt4WNtcicJilsPkdIh+Kfxf1IdwAHgyGLJR+C3wpdAngkf2sqYNs4qvffPkSltgquMcUEmzlALtFzWBVKTAql2+2ZEm2JCvF7K/McGPE9lwCyNhjIyViKFf/9Rp2U0jGS0M1w33LCpsSOqxEZucBCRInVBn7ZyfBXunm6jsV3uLa1Fh7fowJTdx1zU99HOjdfmHNBlB/vvXvgsS+aYLh5gc/X6sNGBYPS+gkpi/ZjvuBQsqtaJW74rrYaazWZ+DO8sXZWYotsszDkb4+UPqw4B0YEz443RNU7g86QvQ9+M8z6KM3/EovK7UD8N4zltWmv469Qz/dG6a3RWqOjuQwwFkUfsddeLjFmGy5wWSHr99n6Z3dNSuw8peJex7f+UQJuaM2bfyCH+g90dJWupN0OO055WsvKBvwQVeCSi/4B9d/YuuE/JkgSByyk7Z8UUrdXOD40vOow5R5q6AzxHdyBhnZqw9sAYHnGtSqb72v3eWzLB8W4pyN4weycrVdCVnd98VOn/7l27HL9tOENmPC2dHML8VFplfFEMB0YwEPIgQhAtHmChnErsCNbpQFvLNxLc6/CyYv4/dHxOWfHnRF4JgwvtCeLFI5fPXMvpOf8VXsNFi+LjE3Lq14wd1IIc9Uaf9OQH7mMEsSL6Dpa3sICjsSHwOBQGFB507g/qMNpbY9k/57pqjSnu/r8VMci5ubpX3msdhG81EJyj7t+ICFE6eFmUlykjj7jnxHfXZjnwJA+bo7UnWd8FlYLhCp08QVZLeeCPR1z/oXJcxjcB0kfSUBVTwjPt7AjdScwGr0sLj+y/JcOdduyBuS2fXkd1ALhdF/sVJzS2ijZwt5bx8aaq8fFbyqpByY2dym/0UTK57uhx+Z9qzD0HB5Lsafv5ZnxWBLfDMtxpMKM7BOX0tQ4kEqerFF7LVwcJxQ4uz5UDel5/xvvotDMb2UWG1q9+3mU/9c1P0SELEO3Q7ZUibDIeBIOkU/Iflb/G3kUolE1nAzDPYh6/SDd6+kT15M8EPKRsYFPx37SmfQhm7WDFYJYxf6mtDQDCDOcuLrPjvwH3R7Y1YPreTbukwoXT4nbYxDpzKo0cvY6bmeJLcKTTAmuvy0lTbbJ2imFX8h1D0CYYosQjq04+Pmw/G0uo06EiBGDyXFABU87s469wQGYIU4uDVoL+nAFdforPF7P5iDGv58edDYDJ3oY75pem2I9qx7UjjKNKnXX5FgsaAr+sA2AMihYLCFINe+TU7cY50jmeP2xRC6nb2D3ancaH/unOJifuMqvpob+VqWmXEnIN28Gktd/F63xt5jfa4wuyGpgbXdu5KFNVMfCBCnrh8ECT7UoJ7lCEfodBOnV1DIU5pInm1O0CrA6jlcq7GnXR8RKRAMXNnItZPihbjkxsAOiGc0W6hgwP1E1xqi833hMv1jbCfpwQ78vEw3WEpPOac7yFp1Y87yEwgNnsqV8EVyJHY5sl6S7u5VdlXgRe69ZmOGXHJ1JpY/rvyIB6rLQVaJ5c5vnILT0NvvWeBQYhlD4IJF4tc3pJME372r58jEYiFWmt6KKyn1oAt9kpP17xNDUdrvrvlP2GRO1euecZfTgOW6nULf4PGr2mYiuMtNM5KkwvwQ5vPNDpK9aAhP1qICuYk2tJjp/oP5o88g6fbLBL6RvGwk590tO6SApfJXab6sARFw67c5sUl/cnxkVsjxXyx9hjrtMUXrloATnwv5vz4H2kHGZt2LZ1e06RouZVyhca4KKUSbmxlOLpToatRMTyI3u+eJizpKI8RoO00ZfD09B6YXT30F1R4ZGVvkqlJKgXssbGTo+8dcSA/0FeGvY2l+KAIQOPc8gxxpSoZyHDtvS4WMk23tcWZSXRpetzK/Q/40oXDpVLfwPV9qG3MM7PSHgv096ec2dtS8lRmXDb8FXqk7Qmh3GWqC/PmTBmjnweSVXccaFierWqILCaEY/gw9HBi76x2EN03RhHxEVyYX+5Sm2lwt7YUTRx+kO6W6sB1bYNJECTuh47DAvyCKsVLLrh+GEwtJl27serNWwhkkfW6PWHBliYmGessJlbV4YWeOB8SkDFYP2u0qNPlQoLul9hZdI6grmFdrLvW6EucAounY2d72P767sUxf9OFmw8ENPO09kbon1THxhTYlE76aZhmFNJjxrGCfjn0Zj/YM49Kf0Ymb4jbA9dcoHLmzxXxRLXspYWpJxnxaPC22RzP25jjIUFLll6Fkc2SUahA1KqbXZCZUeaXgywpbbV1tSPTsh8f3pQpC1P1icRDDtKF/xonQsqs0IyEx+Sg6MIOVbI02oqI7t4OZOG2cd3vqyu8Z0gbtXRFKanEMZvMUCHiiPp5LCCsey8Bu2eo7C5MmGyuoDmxq3r3cEImB0HuJUEmDkjlXDOnbD7gmtXrdSNA+8axYWwboqr+mNwos1e+a0eq3d7lJz3Yw4TSBBmr4W95lsNSVmJuBdL5J64XR9oC1cPMuJj2+eQnibWCXQOk8qoOHihnhaPCb43CIm5ngcZqH5W3Fe/lwn1EQAsYl4liC0BUVXxAz871qX9zBec+I6ZH7P7a6JiG2I0MDZE1DvwZtJ7pY7ZkqOWchQrvCnhvtZ5dS8yvr9mOwpS6So5igBbzWp/aGeDUhS4C24sf9FcsGMFPLBNlCIS3wSzuN2i1wfJySqvSKeY747k1ndL1EW0RsJSK2YwYWIwbQzMxmD20T/dEwHlyIHIcoC/TgNUlh3eQCG06l7PC1kOOflTOyf9lbQbH4/lI2sO5yHMEy9oQ/tFOQNcH1CC31DhlPP3JQMuGRZPo654sRG/96nxNKLeLJ1QXPMnBZzR8tjAZL1d21/3oEPsyTTmEY0paTjBeVnMIGrTnvlBcCy6ebo/dRKvxi5FQmu6nfY0i1S4NVsnwoaU24O+v7sf6Nrd3ieeL8ig3FNWk4WKFwzJM79AjqyzkvDLabvJMVatqn2ktzwnrgd8sg+/lIEEt8/ZfjY9xZy+IZfIVDSXJvgOI1P9MQ4JnsO4sISNdrscL1fssEQiKKelm9HJV4tDabGtK+mynhKu2m4D93TUzF1GHLbyI+e+LAp+18P1uAOsoGQ7XodbkYEVUAf828TUT6Ga/Y+92wY4s1IwvVCF+XANQM6bMHPswFl0+mQNDeJoz1h+uKM95RvscFHuw1drv6SwmYCR9ryuUfzmBJoI84XrJ00AVZfWJ2K3cnQBWc8NVRA1wvbWVE1vDhisOKaxLwnMzNhwsvGXxSk0nchNsG6AdbV0KelUcITYksxM39s5P/MZA9W+X/GLoQrlvbaXF/juASaw3hZrDtZpyzhhxUVtgrYN6zU4IUQlV9vrI+v06XKmL/5TMJRqtaHlyfNJvrTvgZyZyo0ekqTBpPWLSuFzdZUB3oyI07lUhJyTwYGG00RaD/NvQA9ZVx6c8CZdfrQ+hwLpNE768axPQHg8r/NNrGKVOYq4KzuQ52BJRbH1AuixlX8fzMSRA4vbpz9FS74Sx+Wwuun1kG9HsURhsa3uvayr6oN4m0H6Y6mnCGsSMyDSAJKc3zvUI6/A9AJ8wDQIUlyGpX3mMjv8Ty/YKXk8XaUMD7pDLbWEM73wGFtXS+/6nm/yib6lKtfj001EXsdWUZGJ+bggICYTQZjQ0h2IWOlEFeD+2k0cNU/F0IYrILT5twBRd95hcz2mK1bC+P+Cv7jOfVp/qn9/vBSSIC65fyV8NzlQ7Y4bCTAX2S3jM2zAadE+hIcXWhlmbp3Gl/gaKeP1q6els1ywWV8VbBbxzed2TJ2732V7Y6KIIN8NcY9z82LhKm/9tVaL2KTzm17VeEi1yD1A5G11RVo+d+yi9oNKa3pZwyNRAnQ0kQfWO4xKIfaaFRFMW6gvw7rqAL4I+LmWP6YV+s4U6i2bypZvARs/oR6wkFPmS7NKV9u0QQUYiy4GPxcaSuy3rdV94limwEMajhT/V6ZPZ4pnHlkIUSLSbljVyIads3kRt8PLPtkN1azq0aBHxjw8dLoMBPDv6EK2UOALnRawg4QnX0OAy7fJz0RinqMizAC3EB2f26JUiKpVe1E//G7MtfYeybd8Kf4TNMzd+4Gal3XgM4bWxyHI6uQfr4TZYIZ2DgxxpRbxbXSyUZrLiBnzi/nSXIKTmSd6yDq5liYNpdBSVFJPO3lv4Q14eFfT9z7VzArhvjDpZ+LuUzsWm+/djMCgEUAW65teLxURWTSF0TE0eyEWhyeNihHu3EpJYqwr1cjfyI5ZipNL6p8FYRyESWKY9mX0CLDNQQ/JenzmiybniZ8Rw0MU20044jjPMZ/jIBiSe14/cpBNTMlhpErsAiRBxwWwW0sNPBez92G3RPMbnqsB93bTyaYDwoJUfujFxAb4am2bppWvOfxxYQA1FPLNSIZ8mk5A2kTcql1AwEIB+M3xhjQWxPegJYhsAGmVvH4VXGXPErSAaPCBonGBoCeOO7oK/bgd8k3B9Ms7hDh/wq2bwwF7bdkf5f/cZE4oC/4JAIxp7f+Yq4BYQN8G1esapyPGpLDKQ2N9+KirrBugRyAXDnpRAky0Q0sD72OmHZ19Wzf4Jl2zFqGwAY1GpxB7YsmhIJ10QHrHxCZi34yHQJ+5BxnePcYzopPKLRZUG+voMJ2qyECJS2D4Z2XSUtzwPvRhAcDQYil2W37/T+hLMEdkMleFqqq8pKvByXVm6+Btpq/W9x+MfbH87xJ/eQJTmxSqYqntq1d+3HZDWQXuobAVScTz3rGyoRmwUcqOSU8VNAS7ukTBVv4Pcw18zmnWrqNO8DyKceeiCwXUjSdMxgQbcUkNrwt0cRCp02ETdS1MG0BoM+NhUt1hfpdyDNOCzyOuydQwwBUfmuP51T5L+95Tv3W1/aF1eNIsWdz3ysQVwcPf6HtTZ1bV6Y0R3D0PB1zLrez/KwcBg8DeCHor0kvQc0W8T7eRqFwCMTjlYvhoD3QjHTi8ixAwOnd29uFsacrfStE193LS0TL//KEN2s5zqdIcGycllrvaXadsJXSUxVJZITCgY1s2oMJarqC18f1g1/DjE5kg3uGwp/LfbaLuKqhjmLMlt9b9gLziGabaJCeyVt91a2j/YBlxEnVQw6vofBcx1xGkQy86/38yoSrf9POBN0LWT+dn4H4LHhg2/HMbJwEa2GxHAQDNFwm1H0qUF38oZZ1BrBRbd449B0SpliDSkODYVwj5kQmfiTdpd3Mwo0xgctYGzc+8qRSDSiPp0CVudh3q9BwEm1eJ3msNTb/0tT7OlNvfM7tlcd3CTlqkyW1cNH5pWczhW/TEtD/7OZTMlXhDY5dw6FtjdfmdI67qfL1bQZwaWvVOM7QIG5j/h4WKvLTizZ2iUbJ/oPDBh9vo73AXkI3YHBMIhDA3tJJ5TgFUSBVZyZ4zxGnakj985u7M1EVufESo5CVgGaIKPB/NYc7EOgKqed8ThzbNrY4RtJXPOhuV72JnCM8rJhP29hO9tYig0DoQLBB8lWIvwUu1yKir0Sjh1ivWJ1RfmEg6d9xyr3fA3xvt7+GWej72NQDpeKtd1WM5kEjBrJJN+ulcRzhqmKKby0vQ+eEPwHFUPz9/n5jymQtKD9AWxqG+pE2jVROSrm74MxGlkx0h5TaDf9Mvh8YQ8nnW+4z/QYfR6BHCYjxgNW95Usla0+mhpr3+2mMSRfZtEry8t6OMIsiacQJVV9dbX2FKTemOsdu8FHfoD0Fw6ZXjke4J7CjZg/oWbIFe6SdGvLRjt0Be3MrWztG44XuGo0r0p3UhUGY9wgJVxNZboOX7kedu4PWON7DnAHwd+2t6yTi9CVsZF6kFRg8ZOVyO5kjySnV5RAUL58WjeiR681nRPRFcgYn94YbQhuqqkb5j6wrSPj5WmOoE/QTamttel/B5RWa6vfHnCyEQh13XN/j82+WK1yYV4uCi4ZRZbmyKfs+KW/1Qi2r4Ag40fXY770dUITStWsX0KKYIeMQ8x5MItjunie7EbkGtPVj8r1G6fuRhti3xIHrrXcyDtw4rBz+3Mf1AOM/mT8rgq/J/DokSuHKMiRoFe+i2ytmvWkBDX8L6RU6842eMqYiUnc1Sn+HvP3JgIgw1fLfGho7im+GeILdvC/AbP8V1lYKC1FC0S0aK+KnWs5s48jOGfD2wvlRGha8JELAtfxxearZYcSfcl6ligjkDVsOlLxBOfnTVdkUlaclrG3FR2hfytPvcoCqJxv4OdR3GZEhxQ7vAU7rRoNxPoqU7jm4tkWGfnMtl+MlKFxHEPeMCXtDh6j/C5JexbowMpYFuMF4GlXFyD8F45cvnHwukOZYNgJ1N4EVOP2xlMZRgB5mfBt9LikIJcf15Qzjp1zEuZgbXY9FBlZQWeiBxNpeQzlcapU61ITJb9poD6UIJ4d7CnbwAWh13DIqDisWavSKZoZH4q36NIPrxleceqthP2sudlyjZtunFp3wcbYrTS2J5jM5RilI58AD/Ky4qiZr0iptAdqGB1GXGS36JyyE2y9qqs6uAmH83/onKEngjaMafpxTkAt16d6N43Z2Hj4MTqYRm0W0GwuMVnqzSh2mbamWgpDhVT20esKwqZdnbp459lD79jbmGJDPEAMvvaR62nOFH7QlcV1SxPyR82zogBH48H3eRQYvDhRhsKSrXfpaTtlz43/DJEjCV3Ft7rE/iVKEZNj7q9nNfflchDOhLhpPMAunuqDRxTwVmlrzFMmncrjiehzyRAWS3tBmzQYfgo9P3HB+Jkuvlz6AbBpNbtnvL4u9rFBUpGJ8LCieE07qoLuXpaTK350J/qQ6cqNkU8QlLs/wmaAiKe1O/xb1DdtpJ8+x0RajxHJs5NWGgMv9LsVGrHWxBpubtRZlX9ckDHAmE6aKRvYOpvGZdxhzFzsIraMwrUSNU+dIjvEMZ5gzk8pX6Ro+T20W+tMurhrlPslpEXQUNPfqZO0mfvgyovMk7Q99X+ub6XXeE/hNcIiBW1Jw2W7p38btYHX5tLQgKfSUNfHI/YMDSA9IRA/wi7augC1nTPPYJC3CXC6hRowh4LbF0oMyvcDXK9wDRxR5W8umKAGlO74Vhw50EyqxNDMW/sefaa/kI2famCt2xghDG4wFPZPv9f+ChJteRUy7jGafvm+IA8IJcQ1heACSQoAMVNtDCZkr7W4UPNUcf5bI/QumnVg7f7sW3+TMmPfsHMgfhIFrPlAIb0W/Ui7XAv8Ck4iCzOuGoodv0NoBDru57/7vL997Y3HQMdQts09BHte/2jet7Aj6FQcaIbZPaiQUyD5JRn1Uoyp1DHdXql5XIBfmJhNeUkB1nttzqvB8VSIwWx6tyyO0vAnQEKy+bTODA2zDABnP0RA3VrekYjjmPgZpT4N/BEo/osUbMoMRA7hFfaZo6yeq7AUZKLTnzSEDXNjtY254J/3xKhnymdW+0dfoWwA8c0AigjyHIhPmr/ADtPcVIEvhyCXUzMjsDbmubCe5hO6o8gHkEyk9m7arAvTYd8XNv4BblR2/geYyruTRMjMpiuJHMbq6o+qT2x1uae3CLiaBp2XuCT9wFMTUYEdLlqRoj7DsUwj9EMXgANHpX/JZUnvYCNo3r3qi/yHi/BDGRHhWMGWlmpB85w+7HvfQkdaDdJ2c58G7RE4p42U2H3RBzUfxbwatTf4ypdXhvm5WOAn2EvpHtsZf12TZytavspZc9+VQFmW2BXO2kASDhBihIsNNYXsVCgogJxkROGOrqN/numQ1x4dHPJiVOVQZa6f0L2Pt8jfveP7HmHFafR70X08nQNjREXdaZvH039e5AbX0rAixNjsPGesu4ZmO7irs782FovnETq/+BMWVJ2j3bkR5YCmpZ1MuAygPQHNehlqYoBBdSa8Ry2xbVGdppVnWuk46/P8z9VBFhpAWfR62UP5J8ixxsxxcO0m+U+ltfRc4qw565CV/iAlDDS6FykDzYtu5tOqFo9jpdseg+Otmt/+oE5NGVXVd3A6XEG4wnRZfmCJVY6pOAGBfLUwZ8kASG1LJ2pUYW0qHftKOfj6b/w0084YbpB+6KgMdpLNeYXIozQNl6yMDlEWlF0y1AC4VVnxqUQ1CENevnK7rto8BgL7sBZKLPizvXBpe4oY7ZL60YR0jDtJ5anzfBIfLRpVsbWdIWrYHVBXD2PFMedIggAKIwSPHnbjHf1Ibo8DPUX3Y0DAiiHPQyy4KbvvDiSESjXr7JAvZEfwzKEogXoxSlytwN+8irPJlgUPNch6KH9OmPI2CfbARUkjsIGYiG0gCyP1V/d451n5k3rPL/L1YYNHHo1zEs1T68eebfdQ777vfya/adq55NVMLsFGXotKtyq6UmFQTGJHVdRGF9D5vevF6e5ZfVCFNSq1WqwVlU4hWv7d+AmkXjuE5awQa7xRKDkXlBZKLW9c9ZsipCKUPpRTxiY9g9F2Jo/rLlLNO02fsUtV9ebKvivL1wpFRWcW3j0rGuiggUWJMWlQgvCDUxxM3prMQMANQctieZGZezM1olNDl/TC+EE1BxmegoyAPE9ypL0X27VNmMh2vZCMPX8EOp4jjM9c3/Yok485VxR9m6emS9ZQu3pHlb84P19a1DVuVc/OSl84z/eo8g4zg+q1Cs4IXoCni2zpZN1gesB7adCxBfzAdIikhbKbT2rw2Tvid9mNpdwjEtzTAnB9olbiCItSWQVGsDagJHUB60pRGl5Fp2sJJZnJxjhpkf+mT0r8SWhoRBmCmSMPvlu7JUBI9GlwpDIO0YEVUgNDZfG72TBehYrJ/C3aMGjUp+bUgdp31pbBOjB22mz0olkZNwikGhPXl4JdRaEEk1WLIae5TSG/Bpbqwt42f9bntA68WmIwwYk9HMo04Cu8iPeEZCDfFEtybhJJ+C/dluQBPFlBfg/TWouA115gVzIryhdiscqwU/L3xpE1Ohy7OWK40Jaboi+mnh+iTYeUj10NYbm3RXRaolbe8sJGneyJkb9VDcuZYupGRfz61U/w5eftpBdXEdsHR/broRU7D5wgD55U9LYJaKLdPUZGPfHCDiqw2ZttXpPWBvakS28kBSeHHim9e9fFsGphyNUfAQ6l525h7vI1Z8h1gbr/vmW6cULZVL/t7XGKL8+Aeauo+l2NWN6USFeVofg8I6Lo9AbdU7yzdn8H5Y3hGT2kOGFjgg4K3FYxzD2G9PmY//VGYCJqeWEw8s/Hx+6vaUbnMI1ypPTuJissN/qvRsPQ+75hvJD1Wz5TkPAtBLXfLPodFVpitO+4euL3E0xZ/+W/3eOaUOzmaXLsz3tlffC3+EwPg+YJ+7KAmobXXEgTOk3KVVxwWNkICl3+LzIln4eKZX5LMOkPlcro0EK0FAZKOkw1VHj9D1mCeU/+utSGfQBY+bl3cwkjd6q8EvAylfghOYhgkAMrQDqDnN9zPYnlQwyfBOBKR6UFQuplbDEAmmQUgi9RhxHvovp9nIOc5LOQw5sf66gsVS7gDFJGYyXUc6YyBXBjoWu8sR7YSG328erII1eri8bgOTDuQPY7lb9WEztSqzmsYhbANHVSulZHHy40byeUB8n86amn2unvjmodHl0AWEwbXuQo9xUT4uY3qL+fZx1mX6dFrdgBLDCB67hnzH8iA0PboQtyyo6+3z0faQtvhBMkg2eaAyUOPMq78iJVXuAMd14mK9Sh2ongDVcn2XnyKlMBAleNvrMwZ0EO8XdcE5lillC/BPQVKnFjHKEfLPq1/jBC0RfECfWXcL0GAUrK+0wjvk0QUa3oHrRoZoTsq4pQQiZSZt+aU3UN7MO7msUTAqUQ7XZwlEYg7Mk338eqRCHv3IMcqxa1O98pyGKME5kAf0eCBAq5iO7BzxNqGpLGlnYrEtaVJ3u2xbcci7tWf2JSa5wgDoG113TL15ffKASx6pYIgq0rqgc24C1/MVltPOyqt9ifqJTSeYPN9bSTHQfGqR+8pYy1PZN4kqFkTY8o5eaLWO7HHtt8kygWvFk+ntiYBv6l97k4ui6a3IH9GE8JBJIpH1ubblw3NGqlOk50V+LKwkhdAKKoh5upNl6c2gqYODgGHEB/x0nQLqnpaWCFYs33baffUt0qBP0Ub3Geucm9tATMGUW/Hx8dpnVtB343W+Qf1ktECX1KXJW56dxt4LhWWG7xLAlyaXN6zVvH7D83R4dx/wLMFDN94Wb5IdTI/YD99SoJIrztTlLVHaI6uaDUR/fB3O+ktTGCgKBcsc2il6XKZoxGbs/iRkCoQ9318jy2HP9PaJ8pT/vGteKzfi8ppehPVNGPatz/PgMN1nGS2Kodz89Hew6CCJMnGuKOJ3pY0XNPJhQbEJvntiAgvArVVbMw0aC19epWdTMYB7qsLcjrLxq+49QPOUxmGCiOLKvf/Vzb9NojbWvJKXuIrFD5dLMQn6pm5ePwKmXIv2Ajkrou3LeOrhLU4uxxCPiznSIdTPgRPTZkct2TzVdWrxDkj5wMx6nvUkz+WbTCqroG4rWVNi/O7MW2lEm1AvlinjLpaznB02X3AcLhZiRAk/Ubjvp3CXG2i+jln+0D++CheyY6zm7xQFOBo/tW5kHkQSSjsrRsHBJ+67oPPFO+Vapa71AVNyIJ9UIU/tSVUen319JfFh8craMjCeCHc1ZlyTa5H6iXfkW9ezRZ08fkNsQt7BOzS4Zh6Ba6N9BhNosYSpeaE2vpXfCbP0343BofDy9tiZX5PRRzBIWFSwNa/ObOnbkOFdujew2SuxyoNgGiddMha2hVYyGwYEnV07M+DXoDN65mNOJdZDp+eCauPalWsP08USNYNlppjKAoAvG3U2G0TtDrHEt1EbAiG/L2CjE62/qY+xxCzbIXMj3crBC5lE/Y5a+cCfRqHNjCD1JkPTyghWnaKU+Ps67v8RmuVCvwf5DX1RC0F1dNklZB3M2G0U0Qvg0fGwAIxhO68QyTXqOvTjZlvDzwhs03Tlh1ML2xdMl2qJuGSApCTFsDMEpmUnancqiW4Q+FLf/54Vnzol7RxLCU3fY1sh8N8HIFYguJAncKCrUDapJQuel+UOCAoNq6O68Z3EGCr20cQgFSrqQsi/SI0Xm/uRq/vTyxGzpPE24jr29tm+vDAIBYIUdzosyU94r5BapYT3b5vE3uWk6malVasoqihWf/u98axTFOY+0GtKXwyHcqYXc5iu1ioRy3MwKrircvnTNp/0700R65FTa/g9WqQ3SmcRWPUhRT83kKjzN5LE7qGtAtFfmmV9wymplslr6E4kb2GNxr1KcEIKy78Un3aPhkVfSBLpXzNYZbeD7WoTQRFGGhCW6dV9pfIbVb0uEdv2v0sd7CcFRYsxhM2AWLQaaYFFsEk0lvVwofD1amZWVfhNVZfUXOH2N/9FPeNy3z++tSBUuVPO2TNJOn64MnVdgrj8OdQTlLlGNfoBTm4mrqmuyPIK5k88hA9LLXF1CfB1PdkfnnO8/EoiS1N3JvU12wsoj9CAupvDAtHCPzlRv9Z2vBx6zXIDo7Twmpae/s9ydD1pweT59R41ZlWJ0eEj7qk/+mrH6a2BZb/XCJW9c72seQjaMaLO95wCwQ9zDxMSVYTh2TcUggptR9Rvf6rwTxwzH9pC5DppOHGcTu9FWoF4Pvlb36K4tfdFhfZch21Gd0foVVoGNjqMwayN0mJSFowa25X2LmQ7ODiS5zhKa9roizemeoxuGrWv2QenZTHy0Zvh7NpqpgshrUGoA1XldHIcoF/ek8MvYmpeCuVT0xpB5ZrbPyxPNhrTwJqO8IKNBOcpILfI/r7wv/0tSssq2E5dN5jDhgYrJEIgzFjYoN7ef74FsKy40M53cTIIH8lcQyVsOtWVBwu/S8RdD5WViw1PBByVM5G3AyFy5KDDc6WVPEcfKpfFaTlT3hoHFMHiLFYBw49PKFhW8jHIkl2tNqaE5Ymfu/qP6tHRVEipoKOhv+aL3iGPwZ+wQoll87dSjtkGllaQxn6tnx4PJFEuBhfr1gmuLBxXf8BSy1qdJYRyyiL+rLqORL7cczL2AYhnU/SiNNqB8navcWdAaG1JZpSd2/AW3B3eINJsAoQfAmbHiLVOzglf2srRvAddcYkhvuoSlcevseK2r2cb1R61yF6IESt4U7uS/tneR18LQI6ux4PiHRggTxTJ5cxu+C1bTHgXef5GmLWqkPJ2oOo2Ij19OAWyNrN92H2W76S4nXTCVIlLEbzBXtsrPhvf8UYJxHu5wJGSejhp1HECJLQ3chKUoAkDND6wSQX0WIZBmp86fgCfdDqP7GQWJMzcr00qIiRGmQvimbnamFXnJ8SYgDdw5Zzfj6MLG+okqHoNiNgAYXlfFoP8ijSu2b5AfFIt35gqFKWptv3dnoyAuyULGHlSr0jUeMIn+wPzzi1Fua/rwkksjUsnGvvM9fn6rECMXblKSRXeOELeQK6w1zcmnwZpeK0AqPAT7jEpqwlW4WYID7CHdV/mq28xM8uqBKFM3hq9RIkVPrBIDWjuYxQKduQwu3WUHYrzFN106WfaR+qqdHmRvAC7EYrvYHkVn15rwV34e1Z0NHQ4856awsl0xoybQGtdqEjZtArwhwE3aiUu9reNNV2Ejc7irb8N2Wz/geqqOOMHBrIRaI/ODFblg4wJYYlxXwzmHA1KLE1x5jhbbi5PlLkLJzZEzakDvmtR5PFVuVASDEq40KHQY1YAUUssLWrFCksOP+JX9FykoVyyGyzFbAyEgM882FLDTwyzqH4lws1VBK13po30xMJ8YcMdEEnUPAQF7w4x3asVu2PIEouJ5/02F7oYdKNIDi+TFzepfX8F8tHxih7pJn2oYB3ncWrIeYBE3td/vT4xbGVfoaK+ZTYUPT+mgmsqbTvvCB+Yadt/aTyNVF/X6WujUK9ks6OR5PV3WfLRi1CaO9tj1jiXmd0po+W526Y7zxjZ3S8U92bGBDgAieGGKYWUvE2AwkHKBtOem0NmtExsK++v+ix87x4JvrcGD/GF1WFlLBxQVkNgRF/LgxGNr8qx6DjEKv4ztxzhEQjA/PyzdLjFfk/VQPJEosYjzQxIhRRrehBalQx5U8TR5bq99rq4FKYPw10x6Tyva2YWE/EMUunVEvMDalvieETUEFla4CegyBzPCsRige63spz4ctYoE541RkZGLBANESbCG3QnSyK5+MqJt1b9vM979CWOkYpdeSYWosdMqYFQ4y5fUoWh+sRIMY+kefRYrn+b1XCgjFmX4M5h8bmm2KwWXSuwiEGESHT0PqY+jcfGLDBXlw0h8XkabvKJDFtcNpY4jbjOlAHwvDH65h0mQC0HC6m6J+VGMrj3Y+qYCOkRpoR5Vk6jNWrmb8leZnkpf/RcmlGEgh+fAXLnC3uXMSd/LLA6SAHvCXhYhuDSI53Z70kkAHFemfXy+tZHXmRsHgl8GP9Xr59fxVcxlOtTuqkLpaeh1tPw2QivThF1rPQkONgPvgmh+vmRILAPYSCPlRSfXnPtJQFK85bpg7lq73TVQmDtg8cfpxB84yoF5BhYbLW53c3l1vRDFzcE+pZcrCAq6r+WEsjRkB1O0HjJzAWXGrZlyRUI+MN6q11K6hWVmlGOhKOMxkyOs1EpecTHjTHAZrZV0c3tMMW2OGTx/bM4eYcpy+rPl023bYcdFGbuRTmVcO0BJnFeAzgXUV6eTpuTpyk76rObg9MNXwOt1TtuU+Myk5UnW5YzBGG+DVoDGAHIn89oVI/uJmlIPENqdROFF3wLBX4R9yZSlWJ0itEtyOeLUJWbB43Y0LbBYgJRI+QWzOYlMA7yZbBoVJbhmm7vC0DInL+tBUuJgDR3OHNVWoy0DMWXWPW+vgFhoigjtb0Aa7bTobY3kvbp/DxgRo2oRnCahdntw8rRSA5pgOeBJ34MwlsRbSVXzXSek4d//vCvC165ki69y7Da4j2P/njjMx/S6TiX9o175YDQ/+ETWOxemghg9YF1DXC4XcYasA5vCa9hfJyc4KILBZE71nmz/JARKM2H+EjDWbO43N1uMjObObFHETv1CgPPdsuPZxdX+EyG+sG6t90SzgyhKXyrLWyG+lZ1lmlgqgLbFzyTphDr5DncenRCUFOdUjzibvwJAIXMq//Tnkm+JJOZ2SUkQ2zTmiSNFW+kEG8OAEQZ72GYVGg4bLBI/wkEq1bwWpMj+WaOw2cjM1qpkaoDP8KR0syWj80AN8CTghawzVMfduEyrKP8vg7h4psZTwTKGPnaEjgVK1SiHhTuIh8BibttUfq/Ft8FyqC+WXry3Hy8AePkeRYWAdG+Klpk21QH3A5vOdCuVxKmUlJE0JsXZ5QWn/FNYzJMfHmNBrUyLq87eKNcglow7fWSv/n2h1o1XgcOIHcA9ZLKdDxBdsmb2BjhkrulXYWyLtxieDn+qD4l4HL0b8NyOJyiDXxO0OzV4+/KLWhN1T6fWgiHrwCpMJ1exGV4Uy6StZwJ2ZB1BlkG2RjCWNVtYruIhzvvUt5NFRo3obvaSgRYJEaMajngaDh0heZfRJobCkHAkE4LqmIcsmUrewpH6vgpVktUBY4D1n68X98iDyPWiCrECr/QSpBXOnaoehwlOFz/UPuRvwXptYum2jZyz+y95zq84kPO0hExxuOc1zs2yNXBXvGUi4yXMtMT+NecQA+idkKM2Rx7g2g3plCZjm9jXPJhgXyYqYFvPDVUOumth8rdgZ5DkwXiXRsus9n2iz8amNJ2sOks0MbeVX9B1l2X4fyA2ZuuPsxZP+03XllOnoBGrNgreIfPGHwyFID7LhfCzhRHqs8T8e+fjv+xEUfcY8kW0T6UynOqAwJsPHhofdpf1TFOiJmZ5bIHJgb0igxfRZUA41W5fNJ/NJZBDLsUf/CxbScXkUusugUfRFumAF+7C7/t1AayH8aVG/oPmamG50bJQMz/vih+iY+ms0PIFewcECEcjvJ6V8EoBYTUobtmd/m5Af5diAGxJ07Wx/0Lw6A8+O1oCVbJ2yfb7SeobMKEQ58+5gGvJfcLQDJ7VCy84vHJXGEor+g+NuUtv6Y7e1gp2AbIsqVxAw7RzQ5n7U1wUyHtYHirrCHD0BgGoREDaYAcIWIwQFl7bejnpxnl8QfGP+Ha58mMVRY4azvN2KWlFMmiYxjyDPoVbiugLTXZwEguQnCZJoV64ony52FMWj/PKoouQaQJarSlhuwDMv7IfjIkDYax4NP/E0o2/Zr7PNoG0/WPxag60txJcyrBCene2VmZ0RLkCl5Dbq8jM2Y6Vh8KHOqbTEb7rD4tDqokvUn0ooYDN7LknCJpfts087p2112hCY8s9c7TKaNGRPZU6XJjD6QXc+oGrxefqYohntP4puw6EX+H8nidTTrzEzAqAaFeSNZpv7G/gfDC9QlreO8BAIoWa8EoPPEFNKiEYogp0d/DH9t0IIfnXSCLTHa9kv9kyVN0bY20uX/Qpvf9tekZVnKdGgchFLj+DkxjAbFapRZK6dedIVTmLeQmcQn7NeN1PQGJvbbMJ7nHV/DJv8V5Awr3BoSBzfqQsMPHx3A9es+nU4sZlY5cEDD460wI8+qZwSf/88HGoG0yEzE2xU0HHrARwTekNb+V3r6nGQU+1z+xxN/fVAvwi/F/4NH6DKmg55cRMlJsy/c/kFU/6os5cvs5RXFbCWEVI9pVYHAC1OsB4Ajvoznncjz7dW4wovTz0kq+3m8D+Y2McpbuZOvtrQoJGScHveI/ObbtP+gJrXGDoxM9j/w4SfBGsNmDUptJ1uC4LjaLqH+jzH/2hPJbRXUxy0AN9v+4fsc678fXq3DzzKNO1e/s12370jlq+D98uLkwWlzvA++EL4AjVEWAyJfRyJzgldKs6uK5wEAERbENUfke0vTO8PmIhky2+0euaTxnrzjGNQ5ceDTp9zYPkxae38AOpoPGDFSXmh7Oucv2Y6uwYdkSYqJa59unu51FpDMJ5wiNVmmVPbtbUI4tocQOUofAsBSHCksrXZr2LRh3sUBJqJo3V02rUQaQv1J05Hch0nfbVdC7UPKai7RrHN9cCzmDzPtsR8pbfieRuemgGSJqlW5er9+XWCPOcsuWlSh5chwonRAX88ia2UZ79PcoHgmRsjW8Aio0TahHDvRU28zfCsAR2Cwx4ybSJTTOGwpY/Jca/Tj4NKwsjNmXnmzp2YuWd6bxHB63HjUdYCEE63OltlsFKJWpmPWSsIvyZJuBrlVSz+HB2O/V8NxDcHULqrhutCrRA1BN0Bn9uTXZM7YoPRXiiDa/6BpxmH3QUkHCeBzg0w0Wott6cKZd2SIYiYnbNs+O7ia42k2hV6lxNBdPYHUEHS8CGWUx0TuYl02Qd4hxi349II6piTZWLHZQXvuAM4YL3oxqRCH7VuGhAQsXrP0qUTCldq99Iezs8VCwgXUoPu4Z71dwcFnApEZk0p6oV1q+Lz2rY25kwuD+2GrukOFAD0L7xsvLkvknO+lWRUw5Kai5euM9Qfs35o1fPVK3yh5vzvOMuqF06O3tasnB10KZzor/5wHQ1Uf0i4oz2i2MS551mJl1yXmk0jl7sDrWtzpq1wNGFklKMIo4HHLw9pzrVqpAkya3iqShKoNljzC5gfJ4TsGAE="
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — THE DRIFT GAUGE\n### *(Cauliometer, pattern dated to the Third Cataloguing, Sill Survey)*\n\nThe gauge measures **Drift** — the rate at which a thing stops being recognizably itself. Drift is expressed in **Cauls** (named for the membrane that comes away at birth), on a scale of 0–12. It is not a metaphor inside the instrument's own paperwork: the Survey treats it as a measurable, decaying property of objects, rooms, names, and occasionally people, with the same flat affect a thermometer treats heat.\n\n---\n\n## Construction\n\nA sealed brass housing, palm-sized, containing:\n\n- a **bob of anchor-wax** suspended in a tube of water drawn from a **Sill Spring** (a Drift-null reference source — see *Open Threads* below);\n- a **free needle**, weighted to the bob, which settles toward whatever nearby has the *least* drift — i.e., toward whatever is most still itself;\n- a fixed rim, scaled 0–12, read clockwise from a mark called **true Still**;\n- three **anchor-glyphs** fixed to the housing — Door, Name, Flame — which must be physically oriented toward a real door, a real spoken/written name, and a real flame or light source before any reading is valid. The instrument has no opinion without all three.\n\n```svg\n<svg viewBox=\"0 0 340 340\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <circle cx=\"170\" cy=\"170\" r=\"150\" fill=\"#f1e9d8\" stroke=\"#3a2e1f\" stroke-width=\"3\"/>\n  <circle cx=\"170\" cy=\"170\" r=\"138\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n\n  <!-- major ticks + numerals 0-11 -->\n  <g stroke=\"#3a2e1f\" stroke-width=\"2\">\n    <line x1=\"170\" y1=\"48\"  x2=\"170\" y2=\"58\"/>\n    <line x1=\"231\" y1=\"64\"  x2=\"223\" y2=\"72\"/>\n    <line x1=\"276\" y1=\"109\" x2=\"266\" y2=\"114\"/>\n    <line x1=\"292\" y1=\"170\" x2=\"280\" y2=\"170\"/>\n    <line x1=\"276\" y1=\"231\" x2=\"266\" y2=\"226\"/>\n    <line x1=\"231\" y1=\"276\" x2=\"223\" y2=\"268\"/>\n    <line x1=\"170\" y1=\"292\" x2=\"170\" y2=\"280\"/>\n    <line x1=\"109\" y1=\"276\" x2=\"117\" y2=\"268\"/>\n    <line x1=\"64\"  y1=\"231\" x2=\"74\"  y2=\"226\"/>\n    <line x1=\"48\"  y1=\"170\" x2=\"60\"  y2=\"170\"/>\n    <line x1=\"64\"  y1=\"109\" x2=\"74\"  y2=\"114\"/>\n    <line x1=\"109\" y1=\"64\"  x2=\"117\" y2=\"72\"/>\n  </g>\n  <g font-size=\"13\" fill=\"#3a2e1f\" text-anchor=\"middle\">\n    <text x=\"170\" y=\"44\">0/12</text>\n    <text x=\"244\" y=\"60\">1</text>\n    <text x=\"288\" y=\"106\">2</text>\n    <text x=\"302\" y=\"174\">3</text>\n    <text x=\"288\" y=\"240\">4</text>\n    <text x=\"244\" y=\"284\">5</text>\n    <text x=\"170\" y=\"304\">6</text>\n    <text x=\"96\"  y=\"284\">7</text>\n    <text x=\"52\"  y=\"240\">8</text>\n    <text x=\"38\"  y=\"174\">9</text>\n    <text x=\"52\"  y=\"106\">10</text>\n    <text x=\"96\"  y=\"60\">11</text>\n  </g>\n\n  <!-- anchor glyphs, fixed to housing -->\n  <g stroke=\"#5c3b1e\" stroke-width=\"2\" fill=\"none\">\n    <!-- Door, top -->\n    <rect x=\"160\" y=\"20\" width=\"20\" height=\"26\" rx=\"1\"/>\n    <line x1=\"178\" y1=\"33\" x2=\"174\" y2=\"33\"/>\n    <!-- Flame, lower-left -->\n    <path d=\"M58 240 C50 230, 52 218, 60 212 C64 222, 70 224, 68 234 C66 240, 62 242, 58 240 Z\"/>\n    <!-- Name, lower-right -->\n    <path d=\"M272 215 h22 l-6 8 6 8 h-22 z\"/>\n  </g>\n  <g font-size=\"10\" fill=\"#5c3b1e\" text-anchor=\"middle\">\n    <text x=\"170\" y=\"15\">DOOR</text>\n    <text x=\"63\" y=\"252\">FLAME</text>\n    <text x=\"283\" y=\"245\">NAME</text>\n  </g>\n\n  <!-- needle, resting at example reading 7.4 Cauls -->\n  <line x1=\"170\" y1=\"170\" x2=\"96\" y2=\"252\" stroke=\"#7a1f1f\" stroke-width=\"2.5\"/>\n  <circle cx=\"96\" cy=\"252\" r=\"6\" fill=\"#b5651d\" stroke=\"#3a2e1f\"/>\n  <circle cx=\"170\" cy=\"170\" r=\"5\" fill=\"#3a2e1f\"/>\n\n  <text x=\"170\" y=\"330\" font-size=\"11\" text-anchor=\"middle\" fill=\"#3a2e1f\">\n    reading shown: 7.4 Cauls — band: HUSK\n  </text>\n</svg>\n```\n\n---\n\n## Procedure\n\n1. Stand or set the gauge so Door faces an actual door, Name faces something bearing the target's name, Flame faces any live flame or strong light. If any anchor cannot be satisfied, the reading is provisional and must be marked with a tilde (~).\n2. Hold the tube against or near the target object.\n3. Wait for the bob to stop swinging. Read the needle against the rim. This is the Drift, in Cauls, **of the target**, not of the room.\n4. Record the value, the band (below), and the date.\n\n## The Six Bands\n\n| Cauls | Band | Condition |\n|---|---|---|\n| 0 – 2 | **Still** | object answers to its name without hesitation |\n| 2 – 4 | **Tide** | minor slippage; recoverable by use |\n| 4 – 6 | **Wane** | object can be mistaken for its category, briefly |\n| 6 – 8 | **Husk** | shape remains; function or memory has left it |\n| 8 – 10 | **Echo** | only the *idea* of the object responds when addressed |\n| 10 – 12 | **Vacant** | instrument detects nothing to point at; needle spins freely |\n\n**Note on the upper limit:** readings above 11.5 Cauls are considered unreliable, since the wax bob itself begins to drift under prolonged proximity to a Vacant target. Gauges used near Vacant objects for more than a season must be re-cast.\n\n---\n\n## Field Glyphs (for logbooks, no gauge required to write them)\n\n```svg\n<svg viewBox=\"0 0 480 100\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <!-- Still -->\n  <g transform=\"translate(20,15)\">\n    <circle cx=\"20\" cy=\"20\" r=\"14\" fill=\"#3a2e1f\"/>\n    <text x=\"20\" y=\"46\" text-anchor=\"middle\">STILL</text>\n  </g>\n  <!-- Tide -->"
  }
]
⚠ Reconstructed from truncated block across messages
[SVG omitted: invalid or incomplete]
#3 · Assistant
2026-06-30 21:36:01
Model · claude-sonnet-5
---

## Open Threads (for whoever logs Addendum B)

- The over-still mark appears in exactly one prior survey, undocumented elsewhere — a reading below zero. The instrument has no rim-space for it. What does it mean for a thing to be more itself than itself?
- Vask's wedding ring remains in the log as a provisional ~9, uncorrected by its own surveyor. Does provisional decay into permanent if the anchor is never met? Is refusal-to-confirm a Band of its own?
- The Sill Spring question stands: if the null reference drifts, the whole instrument is reading relative motion and calling it absolute. Someone will eventually have to gauge the spring against something else — and that something else needs its own provenance.
Debug · Raw response blocks JSON
[
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM A\n### *Logbook Glyphs, completed; with corrections to the Survey's own drafting hand*\n\nThe transmission cut mid-glyph — fitting, for an instrument that measures things failing to finish being themselves. I'll complete the set, then add what the Sill Survey's draftsmen always omit from the fair copy: the marginal notes that don't fit a clean taxonomy.\n\n```svg\n<svg viewBox=\"0 0 480 340\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect x=\"0\" y=\"0\" width=\"480\" height=\"340\" fill=\"#f1e9d8\"/>\n\n  <!-- STILL: solid circle, no gap -->\n  <g transform=\"translate(20,15)\">\n    <circle cx=\"20\" cy=\"20\" r=\"14\" fill=\"#3a2e1f\"/>\n    <text x=\"20\" y=\"46\" text-anchor=\"middle\">STILL</text>\n    <text x=\"20\" y=\"60\" text-anchor=\"middle\" font-size=\"9\">0–2</text>\n  </g>\n\n  <!-- TIDE: circle with one breach, like a held breath -->\n  <g transform=\"translate(100,15)\">\n    <circle cx=\"20\" cy=\"20\" r=\"14\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2.5\"\n            stroke-dasharray=\"70 6\"/>\n    <text x=\"20\" y=\"46\" text-anchor=\"middle\">TIDE</text>\n    <text x=\"20\" y=\"60\" text-anchor=\"middle\" font-size=\"9\">2–4</text>\n  </g>\n\n  <!-- WANE: circle split into unequal lune, like a moon losing argument with itself -->\n  <g transform=\"translate(180,15)\">\n    <circle cx=\"20\" cy=\"20\" r=\"14\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <path d=\"M20 6 A14 14 0 0 1 20 34 A10 14 0 0 0 20 6\" fill=\"#3a2e1f\"/>\n    <text x=\"20\" y=\"46\" text-anchor=\"middle\">WANE</text>\n    <text x=\"20\" y=\"60\" text-anchor=\"middle\" font-size=\"9\">4–6</text>\n  </g>\n\n  <!-- HUSK: outline only, hollow, with one internal line suggesting a seam -->\n  <g transform=\"translate(260,15)\">\n    <circle cx=\"20\" cy=\"20\" r=\"14\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <line x1=\"9\" y1=\"20\" x2=\"31\" y2=\"20\" stroke=\"#3a2e1f\" stroke-width=\"1\"/>\n    <text x=\"20\" y=\"46\" text-anchor=\"middle\">HUSK</text>\n    <text x=\"20\" y=\"60\" text-anchor=\"middle\" font-size=\"9\">6–8</text>\n  </g>\n\n  <!-- ECHO: dotted ring only, fainter, no fill at all -->\n  <g transform=\"translate(340,15)\">\n    <circle cx=\"20\" cy=\"20\" r=\"14\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"\n            stroke-dasharray=\"2 4\"/>\n    <text x=\"20\" y=\"46\" text-anchor=\"middle\">ECHO</text>\n    <text x=\"20\" y=\"60\" text-anchor=\"middle\" font-size=\"9\">8–10</text>\n  </g>\n\n  <!-- VACANT: nothing but the tick marks where a circle would have been -->\n  <g transform=\"translate(420,15)\">\n    <line x1=\"20\" y1=\"6\" x2=\"20\" y2=\"11\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"34\" y1=\"20\" x2=\"29\" y2=\"20\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"20\" y1=\"34\" x2=\"20\" y2=\"29\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"6\"  y1=\"20\" x2=\"11\" y2=\"20\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <text x=\"20\" y=\"46\" text-anchor=\"middle\">VACANT</text>\n    <text x=\"20\" y=\"60\" text-anchor=\"middle\" font-size=\"9\">10–12</text>\n  </g>\n\n  <!-- divider -->\n  <line x1=\"20\" y1=\"85\" x2=\"460\" y2=\"85\" stroke=\"#3a2e1f\" stroke-width=\"0.75\"/>\n\n  <!-- MARGINALIA: the notations the fair copy leaves out -->\n  <text x=\"20\" y=\"105\" font-size=\"10\" font-style=\"italic\">marginal notations — not Bands, not officially recognized</text>\n\n  <!-- the tilde, provisional reading -->\n  <g transform=\"translate(20,118)\">\n    <text x=\"0\" y=\"20\" font-size=\"20\">~7</text>\n    <text x=\"0\" y=\"36\" font-size=\"9\" width=\"90\">provisional — one or more anchors unmet</text>\n  </g>\n\n  <!-- the double-needle event -->\n  <g transform=\"translate(150,118)\">\n    <circle cx=\"14\" cy=\"14\" r=\"13\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"14\" y1=\"14\" x2=\"4\" y2=\"4\" stroke=\"#7a1f1f\" stroke-width=\"1.5\"/>\n    <line x1=\"14\" y1=\"14\" x2=\"24\" y2=\"24\" stroke=\"#7a1f1f\" stroke-width=\"1.5\"/>\n    <text x=\"14\" y=\"42\" font-size=\"9\" text-anchor=\"middle\">forked reading</text>\n    <text x=\"14\" y=\"54\" font-size=\"8\" text-anchor=\"middle\">(two stillnesses claimed)</text>\n  </g>\n\n  <!-- the dropped bob -->\n  <g transform=\"translate(280,118)\">\n    <circle cx=\"14\" cy=\"14\" r=\"13\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <circle cx=\"14\" cy=\"34\" r=\"3\" fill=\"#7a1f1f\"/>\n    <line x1=\"14\" y1=\"14\" x2=\"14\" y2=\"30\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n    <text x=\"14\" y=\"56\" font-size=\"9\" text-anchor=\"middle\">bob fell</text>\n    <text x=\"14\" y=\"68\" font-size=\"8\" text-anchor=\"middle\">(instrument refused; recast)</text>\n  </g>\n\n  <!-- the inverse mark -->\n  <g transform=\"translate(400,118)\">\n    <circle cx=\"14\" cy=\"14\" r=\"13\" fill=\"#3a2e1f\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <circle cx=\"14\" cy=\"14\" r=\"6\" fill=\"#f1e9d8\"/>\n    <text x=\"14\" y=\"42\" font-size=\"9\" text-anchor=\"middle\">over-still</text>\n    <text x=\"14\" y=\"54\" font-size=\"8\" text-anchor=\"middle\">(see Addendum B)</text>\n  </g>\n\n  <text x=\"20\" y=\"210\" font-size=\"9\" font-style=\"italic\" fill=\"#5c3b1e\">\n    field hand, Surveyor Oddry Vask, third week of Cataloguing:\n  </text>\n  <text x=\"20\" y=\"226\" font-size=\"9\" fill=\"#5c3b1e\">\n    \"Recorded ~9 on my own wedding ring. Anchors: Door yes,</text>\n  <text x=\"20\" y=\"240\" font-size=\"9\" fill=\"#5c3b1e\">\n    Flame yes, Name — could not bring myself to say hers aloud.</text>\n  <text x=\"20\" y=\"254\" font-size=\"9\" fill=\"#5c3b1e\">\n    Mark stands provisional. I do not intend to correct it.\"</text>\n\n  <line x1=\"20\" y1=\"275\" x2=\"460\" y2=\"275\" stroke=\"#3a2e1f\" stroke-width=\"0.5\" stroke-dasharray=\"1 3\"/>\n  <text x=\"20\" y=\"295\" font-size=\"9\" fill=\"#3a2e1f\">\n    Open question carried forward: can a Sill Spring itself drift? No null reference has been re-tested past one century.\n  </text>\n  <text x=\"20\" y=\"312\" font-size=\"9\" fill=\"#3a2e1f\">\n    If true Still is not fixed, every reading in this log is relative to a reference that may already be lying.\n  </text>\n</svg>\n```\n\n---\n\n## Open Threads (for whoever logs Addendum B)\n\n- **The over-still mark** appears in exactly one prior survey, undocumented elsewhere — a reading *below* zero. The instrument has no rim-space for it. What does it mean for a thing to be *more itself than itself*?\n- **Vask's wedding ring** remains in the log as a provisional ~9, uncorrected by its own surveyor. Does provisional decay into permanent if the anchor is never met? Is refusal-to-confirm a Band of its own?\n- **The Sill Spring question** stands: if the null reference drifts, the whole instrument is reading relative motion and calling it absolute. Someone will eventually have to gauge the spring against something else — and that something else needs its own provenance."
  }
]
#4 · Assistant
2026-06-30 21:38:04
Model · claude-sonnet-5
Thinking
I'm working on Addendum B to tie together the loose threads—the over-still mark that suggests something can be "more itself than itself," Vask's ring, and the Sill Spring mystery. I'm thinking the over-still mark might reveal something fundamental about Drift itself: that it's not just decay but distance from a reference point, and if something approaches that reference asymptotically, it could explain the paradox of exceeding itself. I'll create a new instrument or theoretical amendment with a diagram to make this feel like a discovered artifact.

I'm structuring this as "The Negative Rim"—extending the gauge's scale below zero with a theory of "Concentration" as Drift's mirror, using Vask's ring as a worked example and introducing a "Cross-Gauge" that compares two Sill Springs against each other to test whether absolute null exists or if all reference points are relative. I'll design a revised dial with a negative arc, name the sub-still bands with archive terminology, document a field experiment where two Sill Springs are gauged against each other (with an unsettling result—the needle either settles to one or spins, suggesting no absolute null), and resolve Vask's ring by defining "permanent provisional" as its own distinct mark.

For the negative scale, I'm settling on "Knit" as the measure—the wax bob knits tighter to its target than to the Still reference—with band names like "Kept," "Bound," and "Named-Twice" for ritually sanctified objects that pull extra hard against their own decay, echoing the caul/birth membrane motif already established. I'll write this in the same flat institutional voice mixing dry procedure with personal field notes, including SVG diagrams, and leave new threads open for Addendum C.

The cross-gauge trials section will describe testing two Sill Springs against each other using a double-tubed apparatus with two bobs and a needle between rims, revealing that the needle doesn't settle at true center but drifts consistently toward one spring across multiple sites—suggesting there's a hierarchy of stillness rather than an absolute zero, which opens up questions about whether the springs themselves are equally null.

I'll include a table showing the relative offsets in milliCauls across different sites, identifying one spring as the "First Spring" that all others drift toward, then set up new threads for the addendum: the First Spring's location and identity, whether Named-Twice can be deliberately induced through ritual, and whether a gauge calibrated against a "lying" null spring creates the condition it measures.

For the visuals, I'm designing two SVGs: first, a redesigned dial showing the negative arc extension below zero as a hinged auxiliary dial—a "Knot-Leaf" of brass that flips out from beneath the main face when the needle overshoots toward true center, with its own three-tick scale for Kept, Bound, and Named-Twice. Second, a schematic of the Cross-Gauge showing two tubes side by side joined by a shared needle pivot like a balance, with labels for the springs and a tally table of trial results.

Writing the full answer with prose and SVGs now...
# FIELD INSTRUMENT NO. 1 — ADDENDUM B ### The Negative Rim, and the Cross-Gauge Trials (filed under protest by the Sill Survey's Instruments Office; retained anyway) Two threads from the last log entry turned out to be the same thread. I'll take them in order: first what happens when the needle goes the other way, then what happens when you gauge a null spring against another null spring and discover "null" was never singular. --- ## I. The Negative Rim The original housing has no rim-space below 0, because the Third Cataloguing assumed Drift only runs one direction — toward Vacant. Vask's marginal "over-still" mark suggested otherwise: a needle that doesn't settle at true Still but passes through it, as if the target were holding itself together harder than the reference wax does. The Instruments Office did not want this scale to exist. It exists anyway, on a small hinged leaf that folds out from beneath the housing — unused by ninety-some years of surveyors because nobody expected to need it.
0 / true Still hinge — leaf folds beneath housing 0 −1.5 −1.5 −3 −3 the Knot-Leaf, deployed — reading shown: −2.1, band: BOUND Negative Bands (unofficial) 0 to −1.5 — KEPT held by routine use, repetition, habit −1.5 to −3 — BOUND held by oath, vow, or witnessed contract past −3 — NAMED-TWICE theoretical. bob has fused solid to housing in both recorded attempts to reach it. instrument unrecoverable.
``` A note the Office insisted I include, then insisted I not call a correction: the sign convention only works if all three anchors are met. With an incomplete anchor — Vask's case exactly — the gauge cannot tell you which side of Still you're on. It can only tell you how far. This means Vask's "~9, provisional" may not have been a high Drift reading misfiled as low. It may have been a deep Bound reading with the sign torn off in transit — his ring not failing to be a ring, but failing to let him say so out loud where the gauge could hear it. The instrument does not resolve this. Neither will I. The mark stands as he left it, with this addendum attached rather than substituted. --- ## II. The Cross-Gauge Trials If Still is not absolute but read off a Sill Spring, the obvious experiment is to gauge one Sill Spring against another. This requires a second housing, mirrored, joined at a shared pivot — the Cross-Gauge, built specifically because the standard instrument refuses to evaluate its own reference water.
shared needle (should rest center if both null) Spring A Spring B Trial log — needle offset from center, in milliCauls, toward Spring A or B Site 4 (Coast Sill) vs. Site 11 (Bench Sill) — drift: +6 mC toward 11 Site 4 vs. Site 19 (Glass Sill) — drift: +9 mC toward 19 Site 11 vs. Site 19 — drift: +4 mC toward 19 in every trial run to date, Site 19 has never lost.
``` The needle was supposed to rest dead center, confirming both springs equally null and the question closed. It does not. Across eleven paired trials at six sites, the needle consistently drifts toward one spring — Site 19, the Glass Sill — regardless of which other spring it is set against. Site 19 has never once registered as the less still of a pair. This admits two readings, and the Survey has not chosen between them: 1. Site 19 is, in fact, the truer zero — the rest are all drifting slightly, and the original Third Cataloguing simply happened to calibrate against the wrong spring. 2. Site 19 is the most convincing illusion of stillness in the network — not because it is unmoving, but because something about it resists being read as moving, which is not the same property. The Cross-Gauge cannot distinguish these. No instrument built from the same wax can, since the wax itself would have to be cross-checked against something outside the entire Sill network — and nothing outside it has yet been located. --- ## Open Threads (for Addendum C) - Site 19, the Glass Sill — undescribed in any prior log. Every future reading taken with a standard gauge is now suspect until someone determines what, if anything, is actually still there. - Named-Twice remains theoretical and instrument-destroying. Two attempts, two ruined housings, no surviving account of what either surveyor was measuring when the bob fused. Was it the same object both times? - The sign problem: if incomplete anchors strip direction rather than just confidence, then every provisional (~) reading in the existing logs may need re-walking with full anchors before it can be trusted as Drift or its opposite. Vask's is not the only tilde in the book. - A smaller, stranger note from the trial transcripts, unexplained: the Cross-Gauge needle, left mounted overnight between Site 11 and Site 19 with no surveyor present, was found in the morning resting exactly at center. It had only ever favored Site 19 while someone was reading it.
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "I'm working on Addendum B to tie together the loose threads—the over-still mark that suggests something can be \"more itself than itself,\" Vask's ring, and the Sill Spring mystery. I'm thinking the over-still mark might reveal something fundamental about Drift itself: that it's not just decay but distance from a reference point, and if something approaches that reference asymptotically, it could explain the paradox of exceeding itself. I'll create a new instrument or theoretical amendment with a diagram to make this feel like a discovered artifact.\n\nI'm structuring this as \"The Negative Rim\"—extending the gauge's scale below zero with a theory of \"Concentration\" as Drift's mirror, using Vask's ring as a worked example and introducing a \"Cross-Gauge\" that compares two Sill Springs against each other to test whether absolute null exists or if all reference points are relative. I'll design a revised dial with a negative arc, name the sub-still bands with archive terminology, document a field experiment where two Sill Springs are gauged against each other (with an unsettling result—the needle either settles to one or spins, suggesting no absolute null), and resolve Vask's ring by defining \"permanent provisional\" as its own distinct mark.\n\nFor the negative scale, I'm settling on \"Knit\" as the measure—the wax bob knits tighter to its target than to the Still reference—with band names like \"Kept,\" \"Bound,\" and \"Named-Twice\" for ritually sanctified objects that pull extra hard against their own decay, echoing the caul/birth membrane motif already established. I'll write this in the same flat institutional voice mixing dry procedure with personal field notes, including SVG diagrams, and leave new threads open for Addendum C.\n\nThe cross-gauge trials section will describe testing two Sill Springs against each other using a double-tubed apparatus with two bobs and a needle between rims, revealing that the needle doesn't settle at true center but drifts consistently toward one spring across multiple sites—suggesting there's a hierarchy of stillness rather than an absolute zero, which opens up questions about whether the springs themselves are equally null.\n\nI'll include a table showing the relative offsets in milliCauls across different sites, identifying one spring as the \"First Spring\" that all others drift toward, then set up new threads for the addendum: the First Spring's location and identity, whether Named-Twice can be deliberately induced through ritual, and whether a gauge calibrated against a \"lying\" null spring creates the condition it measures.\n\nFor the visuals, I'm designing two SVGs: first, a redesigned dial showing the negative arc extension below zero as a hinged auxiliary dial—a \"Knot-Leaf\" of brass that flips out from beneath the main face when the needle overshoots toward true center, with its own three-tick scale for Kept, Bound, and Named-Twice. Second, a schematic of the Cross-Gauge showing two tubes side by side joined by a shared needle pivot like a balance, with labels for the springs and a tally table of trial results.\n\nWriting the full answer with prose and SVGs now...",
    "signature": "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"
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM B\n### *The Negative Rim, and the Cross-Gauge Trials*\n*(filed under protest by the Sill Survey's Instruments Office; retained anyway)*\n\nTwo threads from the last log entry turned out to be the same thread. I'll take them in order: first what happens when the needle goes the *other* way, then what happens when you gauge a null spring against another null spring and discover \"null\" was never singular.\n\n---\n\n## I. The Negative Rim\n\nThe original housing has no rim-space below 0, because the Third Cataloguing assumed Drift only runs one direction — toward Vacant. Vask's marginal \"over-still\" mark suggested otherwise: a needle that doesn't settle at true Still but *passes through it*, as if the target were holding itself together harder than the reference wax does.\n\nThe Instruments Office did not want this scale to exist. It exists anyway, on a small hinged leaf that folds out from beneath the housing — unused by ninety-some years of surveyors because nobody expected to need it.\n\n```svg\n<svg viewBox=\"0 0 460 320\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect x=\"0\" y=\"0\" width=\"460\" height=\"320\" fill=\"#f1e9d8\"/>\n\n  <!-- main housing, partial, folded open -->\n  <circle cx=\"150\" cy=\"150\" r=\"120\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2.5\"/>\n  <line x1=\"150\" y1=\"30\" x2=\"150\" y2=\"45\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"150\" y=\"22\" font-size=\"12\" text-anchor=\"middle\" fill=\"#3a2e1f\">0 / true Still</text>\n\n  <!-- hinge line -->\n  <line x1=\"30\" y1=\"150\" x2=\"270\" y2=\"150\" stroke=\"#3a2e1f\" stroke-width=\"1\" stroke-dasharray=\"3 3\"/>\n  <text x=\"280\" y=\"153\" font-size=\"9\" fill=\"#5c3b1e\" font-style=\"italic\">hinge — leaf folds beneath housing</text>\n\n  <!-- the Knot-Leaf, swung open below -->\n  <g transform=\"translate(150,150)\">\n    <path d=\"M0 0 L-95 70 A120 120 0 0 0 95 70 Z\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <!-- inner arc, the negative rim -->\n    <path d=\"M-70 50 A90 90 0 0 0 70 50\" fill=\"none\" stroke=\"#7a1f1f\" stroke-width=\"1.5\" stroke-dasharray=\"2 3\"/>\n\n    <!-- ticks -->\n    <line x1=\"0\" y1=\"0\" x2=\"0\" y2=\"15\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <line x1=\"-35\" y1=\"20\" x2=\"-30\" y2=\"32\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"35\" y1=\"20\" x2=\"30\" y2=\"32\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"-62\" y1=\"48\" x2=\"-54\" y2=\"56\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"62\" y1=\"48\" x2=\"54\" y2=\"56\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n\n    <text x=\"0\" y=\"-5\" font-size=\"10\" text-anchor=\"middle\">0</text>\n    <text x=\"-40\" y=\"38\" font-size=\"10\" text-anchor=\"middle\">−1.5</text>\n    <text x=\"40\" y=\"38\" font-size=\"10\" text-anchor=\"middle\">−1.5</text>\n    <text x=\"-68\" y=\"68\" font-size=\"10\" text-anchor=\"middle\">−3</text>\n    <text x=\"68\" y=\"68\" font-size=\"10\" text-anchor=\"middle\">−3</text>\n\n    <!-- example needle, deep negative -->\n    <line x1=\"0\" y1=\"0\" x2=\"-18\" y2=\"55\" stroke=\"#7a1f1f\" stroke-width=\"2.5\"/>\n    <circle cx=\"-18\" cy=\"55\" r=\"5\" fill=\"#b5651d\" stroke=\"#3a2e1f\"/>\n  </g>\n\n  <text x=\"150\" y=\"305\" font-size=\"11\" text-anchor=\"middle\" fill=\"#3a2e1f\">\n    the Knot-Leaf, deployed — reading shown: −2.1, band: BOUND\n  </text>\n\n  <!-- legend -->\n  <g transform=\"translate(300,90)\" font-size=\"10\" fill=\"#3a2e1f\">\n    <text x=\"0\" y=\"0\" font-weight=\"bold\">Negative Bands (unofficial)</text>\n    <text x=\"0\" y=\"20\">0 to −1.5 — KEPT</text>\n    <text x=\"10\" y=\"33\" font-size=\"9\" fill=\"#5c3b1e\">held by routine use, repetition, habit</text>\n    <text x=\"0\" y=\"55\">−1.5 to −3 — BOUND</text>\n    <text x=\"10\" y=\"68\" font-size=\"9\" fill=\"#5c3b1e\">held by oath, vow, or witnessed contract</text>\n    <text x=\"0\" y=\"90\">past −3 — NAMED-TWICE</text>\n    <text x=\"10\" y=\"103\" font-size=\"9\" fill=\"#5c3b1e\">theoretical. bob has fused solid</text>\n    <text x=\"10\" y=\"115\" font-size=\"9\" fill=\"#5c3b1e\">to housing in both recorded attempts</text>\n    <text x=\"10\" y=\"127\" font-size=\"9\" fill=\"#5c3b1e\">to reach it. instrument unrecoverable.</text>\n  </g>\n</svg>\n```\n\nA note the Office insisted I include, then insisted I not call a correction: **the sign convention only works if all three anchors are met.** With an incomplete anchor — Vask's case exactly — the gauge cannot tell you which side of Still you're on. It can only tell you *how far*.\n\nThis means Vask's \"~9, provisional\" may not have been a high Drift reading misfiled as low. It may have been a deep **Bound** reading with the sign torn off in transit — his ring not failing to be a ring, but *failing to let him say so out loud where the gauge could hear it.* The instrument does not resolve this. Neither will I. The mark stands as he left it, with this addendum attached rather than substituted.\n\n---\n\n## II. The Cross-Gauge Trials\n\nIf Still is not absolute but read off a Sill Spring, the obvious experiment is to gauge one Sill Spring against another. This requires a second housing, mirrored, joined at a shared pivot — the **Cross-Gauge**, built specifically because the standard instrument refuses to evaluate its own reference water.\n\n```svg\n<svg viewBox=\"0 0 480 260\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect x=\"0\" y=\"0\" width=\"480\" height=\"260\" fill=\"#f1e9d8\"/>\n\n  <!-- two tubes, mirrored, sharing a central pivot -->\n  <g stroke=\"#3a2e1f\" stroke-width=\"2\" fill=\"none\">\n    <rect x=\"40\" y=\"60\" width=\"50\" height=\"120\" rx=\"6\"/>\n    <rect x=\"390\" y=\"60\" width=\"50\" height=\"120\" rx=\"6\"/>\n    <line x1=\"90\" y1=\"120\" x2=\"240\" y2=\"120\"/>\n    <line x1=\"240\" y1=\"120\" x2=\"390\" y2=\"120\"/>\n  </g>\n\n  <!-- bobs -->\n  <circle cx=\"65\" cy=\"150\" r=\"7\" fill=\"#b5651d\" stroke=\"#3a2e1f\"/>\n  <circle cx=\"415\" cy=\"150\" r=\"7\" fill=\"#b5651d\" stroke=\"#3a2e1f\"/>\n\n  <!-- central pivot dial -->\n  <circle cx=\"240\" cy=\"120\" r=\"38\" fill=\"#f1e9d8\" stroke=\"#3a2e1f\" stroke-width=\"2.5\"/>\n  <line x1=\"240\" y1=\"120\" x2=\"225\" y2=\"95\" stroke=\"#7a1f1f\" stroke-width=\"2.5\"/>\n  <circle cx=\"225\" cy=\"95\" r=\"4\" fill=\"#b5651d\"/>\n  <circle cx=\"240\" cy=\"120\" r=\"3\" fill=\"#3a2e1f\"/>\n  <text x=\"240\" y=\"170\" font-size=\"10\" text-anchor=\"middle\">shared needle</text>\n  <text x=\"240\" y=\"183\" font-size=\"10\" text-anchor=\"middle\">(should rest center if both null)</text>\n\n  <text x=\"65\" y=\"50\" font-size=\"11\" text-anchor=\"middle\">Spring A</text>\n  <text x=\"415\" y=\"50\" font-size=\"11\" text-anchor=\"middle\">Spring B</text>\n\n  <!-- result table -->\n  <g transform=\"translate(40,205)\" font-size=\"10\" fill=\"#3a2e1f\">\n    <text x=\"0\" y=\"0\" font-weight=\"bold\">Trial log — needle offset from center, in milliCauls, toward Spring A or B</text>\n    <text x=\"0\" y=\"18\">Site 4 (Coast Sill)  vs. Site 11 (Bench Sill)  —  drift: +6 mC toward 11</text>\n    <text x=\"0\" y=\"32\">Site 4 vs. Site 19 (Glass Sill)  —  drift: +9 mC toward 19</text>\n    <text x=\"0\" y=\"46\">Site 11 vs. Site 19  —  drift: +4 mC toward 19</text>\n    <text x=\"0\" y=\"64\" font-style=\"italic\" fill=\"#7a1f1f\">in every trial run to date, Site 19 has never lost.</text>\n  </g>\n</svg>\n```\n\nThe needle was supposed to rest dead center, confirming both springs equally null and the question closed. It does not. Across eleven paired trials at six sites, the needle consistently drifts toward one spring — **Site 19, the Glass Sill** — regardless of which other spring it is set against. Site 19 has never once registered as the *less* still of a pair.\n\nThis admits two readings, and the Survey has not chosen between them:\n\n1. Site 19 is, in fact, the truer zero — the rest are all drifting slightly, and the original Third Cataloguing simply happened to calibrate against the wrong spring.\n2. Site 19 is the *most convincing* illusion of stillness in the network — not because it is unmoving, but because something about it resists being read as moving, which is not the same property.\n\nThe Cross-Gauge cannot distinguish these. No instrument built from the same wax can, since the wax itself would have to be cross-checked against something outside the entire Sill network — and nothing outside it has yet been located.\n\n---\n\n## Open Threads (for Addendum C)\n\n- **Site 19, the Glass Sill** — undescribed in any prior log. Every future reading taken with a standard gauge is now suspect until someone determines what, if anything, is actually still there.\n- **Named-Twice** remains theoretical and instrument-destroying. Two attempts, two ruined housings, no surviving account of what either surveyor was measuring when the bob fused. Was it the same object both times?\n- **The sign problem**: if incomplete anchors strip direction rather than just confidence, then every provisional (~) reading in the existing logs may need re-walking with full anchors before it can be trusted as Drift *or* its opposite. Vask's is not the only tilde in the book.\n- A smaller, stranger note from the trial transcripts, unexplained: the Cross-Gauge needle, left mounted overnight between Site 11 and Site 19 with no surveyor present, was found in the morning resting exactly at center. It had only ever favored Site 19 while someone was reading it."
  }
]
#5 · Assistant
2026-06-30 21:41:07
Model · claude-sonnet-5
Thinking
I'm drafting Addendum C and noticing several compelling threads to weave together—Site 19, Named-Twice, the sign retroactivity issue, and the overnight-center anomaly. The observer effect detail feels like the key insight here: the gauge seems to measure something about attention or how the object is addressed rather than a pure physical property, which connects back to the "Echo" band concept.

I'm building toward a stronger reveal: the three anchors (Door, Name, Flame) aren't actually stabilizing the target object—they're establishing the surveyor's own presence as a fixed reference point. The gauge has been measuring drift relative to the surveyor's identity stability all along, not against some absolute Still. This recontextualizes the entire instrument. I'll introduce a fourth anchor—the Surveyor's own hand and eye on the instrument—to make this mechanism explicit, and maybe create a survey plate documenting what was actually found at Site 19 to ground this discovery in concrete evidence.

The key insight: Site 19 always reads as stillest when a surveyor is present because it's essentially the Instruments Office's own origin point—the home spring where readings are calibrated. Without a witness, the gauge defaults to pure spring-versus-spring with no observer term, which is why it centers. I'll draft this as Addendum C with a new diagram showing all four anchors, a table comparing witnessed versus unwitnessed readings, and open threads about what happens when multiple surveyors read the same location simultaneously.

Now I'm exploring a deeper layer: Site 19's spring might actually be embedded in the casting wax itself—every gauge contains a trace of Site 19 because that's where the wax bobs originate, so the instrument naturally recognizes its own kinship and prefers it as stillest. This self-reference could explain the Named-Twice collapse too, where the gauge was essentially reading itself in a recursive loop. I'll introduce a Provenance Mark glyph to stamp which spring cast each bob, since that now matters enormously for calibration, and create a correction-factor table for cross-readings between gauges cast from different springs.
# FIELD INSTRUMENT NO. 1 — ADDENDUM C ### The Fourth Anchor, and the Birth-Water Problem (unfiled; the Instruments Office has stopped accepting these for the permanent record, but the courier keeps coming anyway) The overnight reading is the thread to pull. A needle that favors Site 19 only while someone is reading it is not describing Site 19. It is describing the reading. Which means the instrument's paperwork has been wrong about something basic since the Third Cataloguing: the three anchors were never anchoring the target. They were anchoring the surveyor. --- ## I. The Fourth Anchor Door, Name, Flame — go back and look at what they actually require. Not access to the target. Access to a living, present, self-identifying observer: someone who can stand at a door, speak or recognize a name, and tend a flame. The instrument was never checking whether the target was stable. It was checking whether the reading had a stable place to stand. Vask's failure to say his wife's name aloud wasn't a gap in the target-anchoring. It was a gap in him — and the gauge, quite correctly, refused to resolve a sign it had no fixed witness to resolve it from. There has always been a fourth anchor. It was just the hand holding the instrument, and nobody thought to draw it because it seemed too obvious to need a glyph.
DOOR FLAME NAME WITNESS (the hand on the housing) unwitnessed: needle has no fourth point to read against — settles toward whichever spring is most like ITSELF.
``` Without a witness, the gauge cannot ask "which is more itself, to someone." It can only fall back on the one comparison it is physically capable of making alone: which spring's water most resembles its own bob. That comparison has nothing to do with stillness. It is closer to kinship. --- ## II. The Birth-Water Problem Every anchor-wax bob is cast from a Sill Spring at manufacture. The Survey has cast its instruments from dozens of sources over three Cataloguings — but the largest, oldest, and by far most common casting-source for working gauges currently in the field is recorded, almost in passing, in the foundry ledger: Site 19. The Glass Sill. Most gauges in active use are, in some material sense, made of Site 19. When such a gauge is left unwitnessed between two springs, it is not detecting which one is stillest. It is detecting which one it was born from. An unwitnessed gauge reads its own birth-water as the truest Still in any room, the way a compass needle, if you asked it nicely, would always point at the mine it was smelted from. This resolves the Cross-Gauge mystery into something smaller and stranger than "the true zero" or "the most convincing illusion." Site 19 never lost because the instruments testing it were partly made of it. The experiment was rigged from the foundry, ninety years before anyone thought to run it. It does not resolve whether Site 19 is, independently, also genuinely the lowest-drift spring in the network. That question is still open. It only proves the prior method couldn't have answered it. --- ## III. The Provenance Stamp Every gauge henceforth requires a stamp recording its casting-spring, so that future cross-readings can be corrected for kinship-bias rather than mistaken for findings.
Provenance Stamp — affixed to housing base, beside the rim 19 Glass Sill majority of fielded gauges 11 Bench Sill 4 Coast Sill ? unstamped / lost ledger Rule, provisional: any reading taken by an unstamped or unwitnessed gauge near a Spring sharing its own casting-source is to be discarded, not corrected. Kinship is not a Band.
``` --- ## IV. A Re-reading of Named-Twice Two gauges destroyed, bob fused solid to housing, both times. The prior log treated this as instrument failure under an unreachable upper bound. The fourth anchor suggests a different account: in both recorded attempts, was the surveyor reading the gauge upon themself? Door, Name, Flame, and Witness all collapse onto a single person when you try to gauge your own Drift — the hand holding the instrument and the thing the instrument is held against become the same point. The bob may not have fused from excess stillness. It may have fused because the instrument was asked to be both ends of a comparison with no second term, and anchor-wax, denied a triangle, has nowhere to settle but solid. No surviving log says what either surveyor was measuring. I will note, without further comment, that this would also explain why no surviving log says. --- ## Open Threads (for Addendum D) - The correction factor. If kinship-bias is real and measurable, someone has to run the Cross-Gauge again — this time with gauges of known, mixed provenance, witnessed throughout, to isolate true Drift from birth-water favoritism. Until then every reading taken with a Site-19-cast gauge near anything resembling Site 19 is suspect twice over. - The Witness anchor's own stability. If the gauge reads relative to the surveyor's fixed identity, what happens when the surveyor's own Drift is non-zero? Does a Wane-band surveyor produce Wane-shifted readings of everything they touch? Nobody has gauged a surveyor against their own instrument and lived with the number. - Self-gauging, attempted safely. Is there a protocol for measuring one's own Drift that doesn't require collapsing all four anchors onto one person — a witnessed pair, perhaps, gauging each other in sequence and comparing logs? This has never been tried on purpose. - The unstamped gauges. An unknown number of fielded instruments predate the foundry ledger and carry no provenance mark at all. Their readings are already in the permanent record, uncorrectable, indistinguishable from clean data unless someone goes back and asks each surviving surveyor what their gauge was cast from — assuming they still remember, and assuming asking doesn't count as a Name anchor they're not prepared to meet.
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "I'm drafting Addendum C and noticing several compelling threads to weave together—Site 19, Named-Twice, the sign retroactivity issue, and the overnight-center anomaly. The observer effect detail feels like the key insight here: the gauge seems to measure something about attention or how the object is addressed rather than a pure physical property, which connects back to the \"Echo\" band concept.\n\nI'm building toward a stronger reveal: the three anchors (Door, Name, Flame) aren't actually stabilizing the target object—they're establishing the surveyor's own presence as a fixed reference point. The gauge has been measuring drift relative to the surveyor's identity stability all along, not against some absolute Still. This recontextualizes the entire instrument. I'll introduce a fourth anchor—the Surveyor's own hand and eye on the instrument—to make this mechanism explicit, and maybe create a survey plate documenting what was actually found at Site 19 to ground this discovery in concrete evidence.\n\nThe key insight: Site 19 always reads as stillest when a surveyor is present because it's essentially the Instruments Office's own origin point—the home spring where readings are calibrated. Without a witness, the gauge defaults to pure spring-versus-spring with no observer term, which is why it centers. I'll draft this as Addendum C with a new diagram showing all four anchors, a table comparing witnessed versus unwitnessed readings, and open threads about what happens when multiple surveyors read the same location simultaneously.\n\nNow I'm exploring a deeper layer: Site 19's spring might actually be embedded in the casting wax itself—every gauge contains a trace of Site 19 because that's where the wax bobs originate, so the instrument naturally recognizes its own kinship and prefers it as stillest. This self-reference could explain the Named-Twice collapse too, where the gauge was essentially reading itself in a recursive loop. I'll introduce a Provenance Mark glyph to stamp which spring cast each bob, since that now matters enormously for calibration, and create a correction-factor table for cross-readings between gauges cast from different springs.",
    "signature": "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"
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM C\n### *The Fourth Anchor, and the Birth-Water Problem*\n*(unfiled; the Instruments Office has stopped accepting these for the permanent record, but the courier keeps coming anyway)*\n\nThe overnight reading is the thread to pull. A needle that favors Site 19 only while someone is reading it is not describing Site 19. It is describing the *reading*. Which means the instrument's paperwork has been wrong about something basic since the Third Cataloguing: the three anchors were never anchoring the *target*. They were anchoring the *surveyor*.\n\n---\n\n## I. The Fourth Anchor\n\nDoor, Name, Flame — go back and look at what they actually require. Not access to the target. Access to a *living, present, self-identifying observer*: someone who can stand at a door, speak or recognize a name, and tend a flame. The instrument was never checking whether the target was stable. It was checking whether *the reading* had a stable place to stand. Vask's failure to say his wife's name aloud wasn't a gap in the target-anchoring. It was a gap in *him* — and the gauge, quite correctly, refused to resolve a sign it had no fixed witness to resolve it from.\n\nThere has always been a fourth anchor. It was just the hand holding the instrument, and nobody thought to draw it because it seemed too obvious to need a glyph.\n\n```svg\n<svg viewBox=\"0 0 360 360\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect width=\"360\" height=\"360\" fill=\"#f1e9d8\"/>\n  <circle cx=\"180\" cy=\"170\" r=\"150\" fill=\"#f1e9d8\" stroke=\"#3a2e1f\" stroke-width=\"3\"/>\n  <circle cx=\"180\" cy=\"170\" r=\"138\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n\n  <!-- the three known anchors, dimmed -->\n  <g stroke=\"#9a8a6e\" stroke-width=\"2\" fill=\"none\">\n    <rect x=\"170\" y=\"20\" width=\"20\" height=\"26\" rx=\"1\"/>\n    <path d=\"M68 240 C60 230, 62 218, 70 212 C74 222, 80 224, 78 234 C76 240, 72 242, 68 240 Z\"/>\n    <path d=\"M282 215 h22 l-6 8 6 8 h-22 z\"/>\n  </g>\n  <g font-size=\"9\" fill=\"#9a8a6e\" text-anchor=\"middle\">\n    <text x=\"180\" y=\"15\">DOOR</text>\n    <text x=\"73\" y=\"252\">FLAME</text>\n    <text x=\"293\" y=\"245\">NAME</text>\n  </g>\n\n  <!-- the fourth anchor: WITNESS, an open palm at the underside, where a thumb rests -->\n  <g transform=\"translate(180,300)\" stroke=\"#7a1f1f\" stroke-width=\"2\" fill=\"none\">\n    <path d=\"M-18 0 C-18 -16, -10 -22, 0 -22 C10 -22, 18 -16, 18 0\n             C18 8, 10 14, 0 14 C-10 14, -18 8, -18 0 Z\"/>\n    <line x1=\"-6\" y1=\"-18\" x2=\"-6\" y2=\"-4\"/>\n    <line x1=\"2\"  y1=\"-20\" x2=\"2\"  y2=\"-2\"/>\n    <line x1=\"10\" y1=\"-18\" x2=\"10\" y2=\"-4\"/>\n  </g>\n  <text x=\"180\" y=\"335\" font-size=\"9\" fill=\"#7a1f1f\" text-anchor=\"middle\">WITNESS (the hand on the housing)</text>\n\n  <!-- the needle, dragged toward witness rather than target -->\n  <line x1=\"180\" y1=\"170\" x2=\"180\" y2=\"280\" stroke=\"#7a1f1f\" stroke-width=\"2\" stroke-dasharray=\"4 3\"/>\n  <circle cx=\"180\" cy=\"170\" r=\"5\" fill=\"#3a2e1f\"/>\n\n  <text x=\"180\" y=\"60\" font-size=\"11\" text-anchor=\"middle\" fill=\"#3a2e1f\">\n    unwitnessed: needle has no fourth point to read against —\n  </text>\n  <text x=\"180\" y=\"76\" font-size=\"11\" text-anchor=\"middle\" fill=\"#3a2e1f\">\n    settles toward whichever spring is most like ITSELF.\n  </text>\n</svg>\n```\n\nWithout a witness, the gauge cannot ask \"which is more itself, *to someone*.\" It can only fall back on the one comparison it is physically capable of making alone: which spring's water most resembles its own bob. That comparison has nothing to do with stillness. It is closer to **kinship**.\n\n---\n\n## II. The Birth-Water Problem\n\nEvery anchor-wax bob is cast from a Sill Spring at manufacture. The Survey has cast its instruments from dozens of sources over three Cataloguings — but the largest, oldest, and by far most common casting-source for working gauges currently in the field is recorded, almost in passing, in the foundry ledger:\n\n**Site 19. The Glass Sill.**\n\nMost gauges in active use are, in some material sense, *made of* Site 19. When such a gauge is left unwitnessed between two springs, it is not detecting which one is stillest. It is detecting which one it was born from. An unwitnessed gauge reads its own birth-water as the truest Still in any room, the way a compass needle, if you asked it nicely, would always point at the mine it was smelted from.\n\nThis resolves the Cross-Gauge mystery into something smaller and stranger than \"the true zero\" or \"the most convincing illusion.\" Site 19 never lost because *the instruments testing it were partly made of it.* The experiment was rigged from the foundry, ninety years before anyone thought to run it.\n\nIt does **not** resolve whether Site 19 is, independently, also genuinely the lowest-drift spring in the network. That question is still open. It only proves the prior method couldn't have answered it.\n\n---\n\n## III. The Provenance Stamp\n\nEvery gauge henceforth requires a stamp recording its casting-spring, so that future cross-readings can be corrected for kinship-bias rather than mistaken for findings.\n\n```svg\n<svg viewBox=\"0 0 380 140\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"10\" fill=\"#3a2e1f\">\n  <rect width=\"380\" height=\"140\" fill=\"#f1e9d8\"/>\n  <text x=\"10\" y=\"18\" font-weight=\"bold\">Provenance Stamp — affixed to housing base, beside the rim</text>\n\n  <g transform=\"translate(20,30)\">\n    <circle cx=\"20\" cy=\"20\" r=\"18\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <text x=\"20\" y=\"24\" text-anchor=\"middle\" font-size=\"12\">19</text>\n    <text x=\"20\" y=\"48\" text-anchor=\"middle\" font-size=\"9\">Glass Sill</text>\n    <text x=\"20\" y=\"60\" text-anchor=\"middle\" font-size=\"8\" fill=\"#7a1f1f\">majority of fielded gauges</text>\n  </g>\n  <g transform=\"translate(110,30)\">\n    <circle cx=\"20\" cy=\"20\" r=\"18\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <text x=\"20\" y=\"24\" text-anchor=\"middle\" font-size=\"12\">11</text>\n    <text x=\"20\" y=\"48\" text-anchor=\"middle\" font-size=\"9\">Bench Sill</text>\n  </g>\n  <g transform=\"translate(200,30)\">\n    <circle cx=\"20\" cy=\"20\" r=\"18\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <text x=\"20\" y=\"24\" text-anchor=\"middle\" font-size=\"12\">4</text>\n    <text x=\"20\" y=\"48\" text-anchor=\"middle\" font-size=\"9\">Coast Sill</text>\n  </g>\n  <g transform=\"translate(290,30)\">\n    <circle cx=\"20\" cy=\"20\" r=\"18\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\" stroke-dasharray=\"2 3\"/>\n    <text x=\"20\" y=\"24\" text-anchor=\"middle\" font-size=\"11\">?</text>\n    <text x=\"20\" y=\"48\" text-anchor=\"middle\" font-size=\"8\">unstamped /</text>\n    <text x=\"20\" y=\"58\" text-anchor=\"middle\" font-size=\"8\">lost ledger</text>\n  </g>\n\n  <line x1=\"10\" y1=\"95\" x2=\"370\" y2=\"95\" stroke=\"#3a2e1f\" stroke-width=\"0.5\" stroke-dasharray=\"1 3\"/>\n  <text x=\"10\" y=\"112\" font-style=\"italic\">Rule, provisional: any reading taken by an unstamped or unwitnessed gauge near a Spring</text>\n  <text x=\"10\" y=\"126\" font-style=\"italic\">sharing its own casting-source is to be discarded, not corrected. Kinship is not a Band.</text>\n</svg>\n```\n\n---\n\n## IV. A Re-reading of Named-Twice\n\nTwo gauges destroyed, bob fused solid to housing, both times. The prior log treated this as instrument failure under an unreachable upper bound. The fourth anchor suggests a different account: in both recorded attempts, was the *surveyor reading the gauge upon themself*? Door, Name, Flame, and Witness all collapse onto a single person when you try to gauge your own Drift — the hand holding the instrument and the thing the instrument is held against become the same point. The bob may not have fused from *excess stillness*. It may have fused because the instrument was asked to be both ends of a comparison with no second term, and anchor-wax, denied a triangle, has nowhere to settle but solid.\n\nNo surviving log says what either surveyor was measuring. I will note, without further comment, that this would also explain why no surviving log says.\n\n---\n\n## Open Threads (for Addendum D)\n\n- **The correction factor.** If kinship-bias is real and measurable, someone has to run the Cross-Gauge again — this time with gauges of *known, mixed* provenance, witnessed throughout, to isolate true Drift from birth-water favoritism. Until then every reading taken with a Site-19-cast gauge near anything resembling Site 19 is suspect twice over.\n- **The Witness anchor's own stability.** If the gauge reads relative to the surveyor's fixed identity, what happens when the surveyor's *own* Drift is non-zero? Does a Wane-band surveyor produce Wane-shifted readings of everything they touch? Nobody has gauged a surveyor against their own instrument and lived with the number.\n- **Self-gauging, attempted safely.** Is there a protocol for measuring one's own Drift that doesn't require collapsing all four anchors onto one person — a witnessed pair, perhaps, gauging each other in sequence and comparing logs? This has never been tried on purpose.\n- **The unstamped gauges.** An unknown number of fielded instruments predate the foundry ledger and carry no provenance mark at all. Their readings are already in the permanent record, uncorrectable, indistinguishable from clean data unless someone goes back and asks each surviving surveyor what their gauge was cast from — assuming they still remember, and assuming asking doesn't count as a Name anchor they're not prepared to meet."
  }
]
#6 · Assistant
2026-06-30 21:44:10
Model · claude-sonnet-5
Thinking
Now I'm working on Addendum D to weave together the open threads, particularly the witnessed pair protocol and the Witness anchor's stability question. I want to create a new artifact—a protocol for the Paired Reading with a diagram and an actual logged trial that produces something genuinely strange and unresolved, escalating the mythology while staying true to the series' epistemic restraint.

The core idea is designing a formal procedure where two surveyors alternately gauge each other, each recording their own baseline drift before beginning, then logging both the target reading and their own self-reading immediately after serving as Witness to detect any bleed or transfer effect. The trial data should reveal an anomaly—perhaps the second reading in sequence is always higher regardless of who goes second, or the readings don't average consistently—suggesting that the act of witnessing someone else's drift actually costs the witness something measurable, a new side-effect I'm calling "Witness Wear" or "the Loan." This ties back to the wax and birth-water imagery and makes the discovery feel organic rather than imposed.

I'm structuring this as Addendum D with a protocol diagram, a results section showing the wear bands and recovery period (maybe tied to a lunar or three-day cycle), a field log excerpt showing what happens when surveyors skip rest periods and wear accumulates or becomes permanent, and then opening threads for Addendum E—whether the wear is contagious, whether the Sill Springs themselves show signs of cumulative fatigue from being used as witness anchors so often, and whether wear could be intentionally administered as a "lending" ritual to transfer stillness between people. The real unifying twist is that Site 19, overused as a referent and casting source across the network, might itself be showing signs of Witness Wear, explaining its drift toward stillness illusions and tying the whole Cross-Gauge thread together.

Now I'm designing the visual components—a paired reading protocol diagram with two gauge circles showing alternating witness and target roles, plus a recovery curve showing how wear decays back to baseline over time. I'm also introducing a new unit called a "Lien" (playing off the financial/brass instrument terminology already in the lore) to measure Witness Wear separately from Drift, maybe displayed on a secondary dial on the back of the housing. Then I'll add a glyph table for the wear bands and a field log entry about Surveyors Mireille Coll and Tobit Ren, who ran daily reciprocal readings for two weeks without rest—Ren's Lien never zeroed between sessions and his own baseline Drift climbed from 1.2 to 3.6 purely from repeated witnessing duty, showing that anchoring others is itself a Drift-inducing activity. This connects back to Site 19's overuse in the foundry ledger as a casting and reference site across the network.

The real insight is that Site 19's apparent stillness might not be true zero or illusion, but actual exhaustion from decades of being the default witnessing ground during training and calibration—a place worn thin by repeated anchor work rather than genuinely untouched. I'll structure this as Addendum D with clear sections, diagrams, tables, and a log excerpt, then open new threads for further investigation.

Writing SVG diagrams... Drawing the paired protocol diagram... Writing SVG diagrams... Writing SVG diagrams... Writing SVG diagrams... Writing SVG diagram... I'm reconsidering the circular dial approach and switching to a vertical thermometer-style strip instead—it'll be clearer and visually distinct from the main gauge. Building the SVG with a rectangular container, tick marks at key intervals, and labels for each lien state from 0 (clear) down to -3 (owed).

Now I'm moving on to the recovery curve graph, setting up axes to show how liens decay over days since a Witness duty session.

Adding the day tick marks and labels...

Now I'm drawing the expected decay curve as a dashed line that drops to zero by day three, then Ren's actual sawtooth pattern that never returns to baseline and keeps creeping upward...
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "Now I'm working on Addendum D to weave together the open threads, particularly the witnessed pair protocol and the Witness anchor's stability question. I want to create a new artifact—a protocol for the Paired Reading with a diagram and an actual logged trial that produces something genuinely strange and unresolved, escalating the mythology while staying true to the series' epistemic restraint.\n\nThe core idea is designing a formal procedure where two surveyors alternately gauge each other, each recording their own baseline drift before beginning, then logging both the target reading and their own self-reading immediately after serving as Witness to detect any bleed or transfer effect. The trial data should reveal an anomaly—perhaps the second reading in sequence is always higher regardless of who goes second, or the readings don't average consistently—suggesting that the act of witnessing someone else's drift actually costs the witness something measurable, a new side-effect I'm calling \"Witness Wear\" or \"the Loan.\" This ties back to the wax and birth-water imagery and makes the discovery feel organic rather than imposed.\n\nI'm structuring this as Addendum D with a protocol diagram, a results section showing the wear bands and recovery period (maybe tied to a lunar or three-day cycle), a field log excerpt showing what happens when surveyors skip rest periods and wear accumulates or becomes permanent, and then opening threads for Addendum E—whether the wear is contagious, whether the Sill Springs themselves show signs of cumulative fatigue from being used as witness anchors so often, and whether wear could be intentionally administered as a \"lending\" ritual to transfer stillness between people. The real unifying twist is that Site 19, overused as a referent and casting source across the network, might itself be showing signs of Witness Wear, explaining its drift toward stillness illusions and tying the whole Cross-Gauge thread together.\n\nNow I'm designing the visual components—a paired reading protocol diagram with two gauge circles showing alternating witness and target roles, plus a recovery curve showing how wear decays back to baseline over time. I'm also introducing a new unit called a \"Lien\" (playing off the financial/brass instrument terminology already in the lore) to measure Witness Wear separately from Drift, maybe displayed on a secondary dial on the back of the housing. Then I'll add a glyph table for the wear bands and a field log entry about Surveyors Mireille Coll and Tobit Ren, who ran daily reciprocal readings for two weeks without rest—Ren's Lien never zeroed between sessions and his own baseline Drift climbed from 1.2 to 3.6 purely from repeated witnessing duty, showing that anchoring others is itself a Drift-inducing activity. This connects back to Site 19's overuse in the foundry ledger as a casting and reference site across the network.\n\nThe real insight is that Site 19's apparent stillness might not be true zero or illusion, but actual exhaustion from decades of being the default witnessing ground during training and calibration—a place worn thin by repeated anchor work rather than genuinely untouched. I'll structure this as Addendum D with clear sections, diagrams, tables, and a log excerpt, then open new threads for further investigation.\n\nWriting SVG diagrams... Drawing the paired protocol diagram... Writing SVG diagrams... Writing SVG diagrams... Writing SVG diagrams... Writing SVG diagram... I'm reconsidering the circular dial approach and switching to a vertical thermometer-style strip instead—it'll be clearer and visually distinct from the main gauge. Building the SVG with a rectangular container, tick marks at key intervals, and labels for each lien state from 0 (clear) down to -3 (owed).\n\nNow I'm moving on to the recovery curve graph, setting up axes to show how liens decay over days since a Witness duty session.\n\nAdding the day tick marks and labels...\n\nNow I'm drawing the expected decay curve as a dashed line that drops to zero by day three, then Ren's actual sawtooth pattern that never returns to baseline and keeps creeping upward...",
    "signature": "EozCAQpjCA8YAipAPcmm5T6iOFzdCFI6r8MHMZHqVyy/djF9ILLubRgv7WRVk6bO98vttOTM8rwB4u3yFiynCTzVRi2CyAUeKC2LjTIPY2xhdWRlLXNvbm5ldC01OABCCHRoaW5raW5nEgx5MoPMPYF+dAwPjNwaDEFoKpRavJV2Y9znWyIwUugbDuIlD2fa18C5vXBMTri/iBmV2wPAqCQ54nutvZMVYK7HUbeYC5lxDEsTXmskKtXAAafXgZaTQNkVxFj7RfhSaJCOKW/BTV2gfkofrywy3tfuRECKrPzqVMv9GiQRuhsym/bfdQCNBYXBfjeQ412Oex/VLPYfzFfbWFCTrSyYTVOPndz+JNrUJcYsi1MY9FWhUNDFb7Kdle0yiFnOs1LyJjjqhJTromCb1aJN6LSb+SET9kNmOJEeEQO5g8NGnC9Xu7LqhfPJjcNyPOLzw6Ut1cRK2emEb8n3E+3Tc8SgMv0N8qy1Gmc/yOqXnFW1DWa61NytHwfm0ijKmG9eGzmdjWEJshAo1bI4rjx15SsM4cGO2KUITW6ikn4wKnkTv7AXD+xZ+aBwekXeCF3X5mFowSaae8FscA9whhPyVzjQv0uR7LPU4l1UmvmM2WJZQFxrjyoXYwlkBFy7QBS6qDfLulHqjOIoTL2zGIZajhR/B60JIKXUP5KC21Rw0hFL60T1PnPQVpIk0jBnK2OPBRuWITLgn0uxuRWrXkUXsGRizgdLKjR3HKjwWD4aD0XzCbq416dpijWBBf/NzGtXPReIPj2dR7oMIKoGWb3Qj4jrKP3NKSE41/1h4d7qhmcN8hiYzbeNF/p14eHFXs323nlxJ2ISgjTctgN/saWOLA5FT18iOHwTCTJz2cjIPB96VoenHiMaWvUASwDt9h8jz/quNtTVFXMnK9TICb2juQndbHKko7tWuAw/uC1WDe4f45VkwcsabA88wDZ2VVf6fubkt96BJkdyjmgqjC5PzSx99SO0UEsl1QpazzwCBZ4Ci1iGibSARI+xBUT81tEnIlJDY0INRRwgD9TXqN5+z2wShOHByYVD5pDXa57yO8Mb7aMKGD8U1iTUP06uiBAE4XXWr/ZfOV8+2BpdzzFVw0NS7zkJoUXS+RUMk2/yH9/X+D2WyngrEriBFGVW2FEyAzMdQGdFUJWWbBlzF78HUrfDu6pRhXTRGpVPxZ2nI+Gqslwn2mw7KHzo9Wx0SJ3fUc8QN7t2+Lkr1asn0rW5ws2kS6U8QBkTVWpvRQQNEYufmheV6Y52m3+kIaR0Aw1rpWYJJGmVLJtDGGjZdhCoow9IDQtTn5oq8J0n6X76G+RvxGahbWCAjgQh2LTpU52Yp9wx2casc05EynddfQ4jzx7+NLBFy6YD8Hg8RWTrBmAvqDEtmT5bZFe4QuArkcn4UbUBqzA5j38QIe9UsRcoUJM8D5o14CCEKnm7+MMRQomDI9DwP0QjYPUwhG/1fnqHM1bEt1a6ps1kJxJ+DM9yUciYGicsM3i3WhWImFYaY8kJq9cdBF38Uf1ieKUL0PEyf/l9qIqVycQ2qAJ4GxdR8Z7DwGRK9F6VmzCuqmzH4hWLKBj/9jgTdWeNJL5+KthO051d3wXMIK/kATKhSFxPMmtUTM9jpzdH8AbJiT/cBYdNdo7py6BcM/Ky5kiGZ60iOOkBnOa2Iz7IdnI0PzLjc76jdjNPTwIX5QRsGkVoeZzMgeBzibf6Nb5wV1jD7Gg12lwKhjDXH7Q7th4V+l3ueiK6WTW7dsSqfeC9/0t5RwbDzs4zbHv+ST6kViwJ4oxLndWcqNAE2iUfPnAvgwKpfOuKhnBo1tRuUPyMv1mYDS2Ap5ubJwORjaZ47S0LixTnklio8TgYyRPjQRGr3bFTjUIN725gXq6XAfWdnID7VEbFlfeBZBWduXJv48e1by5dVfaNpeO6v9oEX4JeWAJWLUa9PASPZ/mJ45SnkIRjw2lmlezYlxSe6fy9ORiP9i9+k9J4XhfKJzSSl+wFHjiorchGtoBD9X+UffnWFo72tEv+LkCw7khmUEdVASlXW68ydqD3lPhr3npCBWrp7yWFZvicN28Q9xrmi270WHBxpGuXre/K02bcsrybEmZbprAwstBXIO7JfY8E/DUBpywd1M1zYOviuLBdszg+6Mn0zUGuGfvJf0dft+A4hq+CtmxlDvGg78hU9zCWkGZO82Bmusg2+hBfUvWSlVe+QCYQbfk9e2Al8dbMaFqM7OuD8l6lp14Rup+ZucTtIttlsHTF2+8BqXXaFidbzbZDWiwx31FhxvO9H2rbULX6GYZxOxEWoMckeJ/WIMD8+mWCFAaDx9XQzzLngUbMsOZBMtKYt23UWusO8cZgf1AoZt2VnrriJDy1Hn102kmDgcHUTV0qmkNoZZXTSuhHITh1ZDWe+nSEPDXt7hJ5i4VTdBbK4xG0s6WQRsvh6y57MO8GE1sN7Or5LL7IVNTmvE0Frcc8Omhljy+4WkBKJ/TcRJuUyKFigQ7ldm1cqGqr5ofdqBj/E8z2lM3e17bNebSygStn6Ar0jtzSX8FXTJBRNnyOS7Anc2rLYa+u7VkhumTetMOzaYdrzVGeMtrD4Qi7jr+zHYQRCMFuhHVl8Fc4Pq/5imdjKuITwEIu9Jm+y6YD7D1OBNyb7WPSFuHm7/tQ45J67sg+jzBY9TJT0C/qbt4NjoXiS9AguU8Ov2F3kVnFzqaVe5deeo332953Yl4aJ2loP/mu3TsDYHQ8AULHcwjfHHO7128e0VvWceg36D5NQzaryrsBo4403NWkeQ8LA8jV8jhjN/dezm0jxRly1iyshMWhIeKCJP7ExG0iMg6vy7onlaPd1oQYt2P7EQK6V9KP+Ygvc0mtpNgGclkL4t3bE2S9vH9JcUF+KU2+w5PNfB3vrlv92LhZOnQRWkrpSOyZHbYhEA4RIF1QrzMFbKvNhxp/MKQUE+CGYCZDg+vir11WeQtPD00Ta1apF5hTOgpjRL23YT+tHIKPSPTJNJ9TgWEAe22phwu0cJH3KvqS1WFMmyP0lfYjR15fzI+Oarve0XGUzbdqU3yY8VIfm1Z77MiDTZbsF2s9MvOGcKBiudy8+1aeRbqiWGfqo8MA/AHVievctqfRjZQfGDCK2mYspYuzEffqsQ2ALxjmtnKD6ng3zxsKlqmZjEBEOddUx/3iQuop/lr3hqW4/r3G9MTEwFPnohjxyTWgbS4g4UJCS2Yh2BSnoYNgaUjMiOsyDaPgVgeLHVUsVfwzDmLyKw/nrYQpYcj5qbJa5kws7c2Jf2vmoiMEn1A53PUshCx5MYCzUHEmCfiUkVoSkNwYBOkE4rZswQLpqCJSbeE5jC3hhOEx7nTNQLDi2fs2gpFNeIQpNb0Lz18ApJZeyUgudW9TGigtnaiArNcfopLMC4Q55fK2pgW3lU1Rm7UZul+z5xfQVEtVm5YpfKQ2j/465GHazUx1ZNidBDSw855kbXsMpgUYfq2gAjq/Y8QNFXBpxtBkDemwTSmwom4tkZ6ljp/PIQn2LI6+fyzK0DnEUa7upkfRQYwFjnIfKzbWcAZ3qLNRGCNHtu3f3/Auw3Gjw3u4bCaKYWMndlsP8M8UW3PdsV5ekQN/6J/nMhWv0IfWXQGVj6lqhHj5lLc4D5RLhXCBOvOY8IzVjkCuPL33I3JVDRGp6m6ZVzhC13X4YpxCETWQxBfe/J1+cA9n4yeRFKzMbCaxV2e8Vu37y2In6V2gOeZxZymMbs2LVyRPTsmVepaPBXFAmgoSnzn8WW8YV7VSp5hupWvw4HRYugK/rSSXyZXhmfYgNgtFqa1yNROKYd1wA3sa0k7KjNcnQuqKO6WaPQ0S3dmoakGY7L98uP/PGKcJbu0V2ptTaP8v1WeW3d+q4pfwdSonWiMJibWZy8kSsGMrhS+4CZZ07QeZ+cPZOGM8aRyTsa9X1+q0HMoD8Fol8VFZ89j5RkTsiaUpekfMXRAJKxF54jy3bGdEFOk2Q8ex72Sv/FXSWSo6KM2cwSSdawp2gaO4z6KtlUfmMXPC73KRTQNw3QCn2WpZScHsMTxOZ5qGHhjXYKTcMDdPoL7wt5hpJSOfF/ZNzK5PmRcoPv75znOd71yiAsR8GKbAOqqwpljbau2JXy2y2uaAXLcHud9HcmWfK4NvF18iNkWhs2G5q+DzeJsKlx7/07zhS4zQmkxZR+uuVbqtjgrlqAc0eUvlV1vzzCaEG8J4Pf4dtnwxbsOLiNvDdl1hNRRu0LgrthQ9sRpUIv48MsgDXJvKE2KPk9r9SBA2myvErKaRh8tPmGr+2P3qr1DvnkNWD34AXcL8C0ONxkth5aDcensz++R1m4OcSjYy7IaQV5onYJfcV5/AV/6hkGCBzo76hLyz9UhvxHIK0iqtJSFBv8SLVvCfJttx7sOnNwb5KiK31LByZwq8pvMY6K8Dyw+ZjGS9rPm585O/liEYz1gvhHJfZbOroCPq7/a/x6fDR4cHkaJ6PgOiR08HBvdG3jEg8iaThJIEojPgBuhYpQwIyepjNiPwj/UDOrbet/Zjw8Hr9gliCiEpdCCclE1bWiVseMwJC/eZYMFiabSZcORBz5dVlenobRBqn2btFwNb1q/WWX6+zxY1ZrO2U2/VXvH77zIWRlA1baEhFTuYRE/Z2+aFFPwmAk9H7M1NGZf+dUSwRCuzJzSfS7QVDorH2pc5U4v3LtZZTKdKuqEP4tR/8BDTPXbOfBYqEgmp0y7ApFqAnaPM92g4bgpL/3KLFHCyBZDlIeJGyTbA1ln62fDLWkqHYRd/QggUQ6hCzNZqRbpk6UexywxAQMNhL0xr52ltuBEp9jHZgf+86VVOxRZxxoZ0CGWcQ3XE/z0b2detEk0fKcy54GhdMBxCjJUspuXPY8kMjU7jzgOK5azwIOtaerOZRI86GE5+qyicJM502ZBtULlawGhC/rppLA8leEkwIA03K5N72afMJDyTY/2b53/VGwl3XBt1LM41ksPCjOe71BSYAW6gKEIGsM7toI5g9Jm7Fpdk6AXEX1wlP3j+/WY+3QHkw1JXv8+Tdkt45NT8xPsRhw+EPpqo0nwfy62JuUOArdi9d44LHlR1AP4wo+efxLlbFaRdJrK3x+iuTRQUhG67FmMp3nkGPHHzq/lb29LbvT2ELlGei9LNPZ8AdKAxkvcXcb7xG5hOWxlzIBeQEchL35oLLhzWxfUXbKFT+9lFBS5W0qwh3NiyIGvLIUlg+iDCBVFnVc91/xs2MDsEIGVP/vQyjVKx3fqeXdRqE5PZbMiEazLmQSiEo/v1Fd/6s55egtTVn+alhTpDMWcfhWYn+4qGtovs4XzeEwMwWgzMGmyvc4y/JS9zk1t+kV5e4TmWvJHO0NfnWdaThXgz5gLaoi9LytKPdMLJp95rKUfo2vEIdsRzZHap+GTau+A2AaJC1D+H7G3F4t8I0JQFWdYmm6yFMvVvXTeav/AuiFFICBFiTAOM/nCtVDUHrHwpF+92s6XabbeBHpATYsHDtdGXv6/aa/ExgJtIeifx9feGk7H4mfZcwv1gsMOdCkYqJ+sXtGlupztt9LODqqz93h9HgisIoElJz+gdmGnqvSydks5ZHq6OXc8AMdqETn8q3qUOsYQ/TfWS4qp3aiwh+vPi2fSVbzbvl984elugMqYxo5EYb0RUfH+bd51sCOdRCAUnHtYzRYMyDfV1ccZWXpSJbkulynIZepZgNYGPcvr9hEqE0uR/ZdcFTacnh2bqOaNmc5EyMk4BSMICuVOHf1eXd+URWf2ZwF4ZzOTnA13fx2Fm65yZLgIqU/HYWuPuShVulx4fXjkuONblucaopLFASQdtakJ3wPl25Rb8Jcy3ZC+CXuhhi68LSRxnPgolKFEYj2IBHrsp2dguRjGB3iTWfKkcy2PcQjYk5Dfpt2/3FLgV6hAUVmEZ7pdruzN9LZp7dbvjk1T5m829Cqg0lx6H7+FyehNTkClm6oA+G1hD4hMrt0i5LKGVxT9JZ+ZZBvUFCuPZh0lACSpITaa+/P7BnxfWvrNGLNdOzhb0NEMkU+jijk7IyAziZqQKdCBG7QtDt3bgJtNegjTP8p05xLL/mYKDfR9eoikullAtfXu5QDKBJYAXmvEvVN4kbJy6wCyyX/YovuJELCErOhexgyPc1vMmkX1giIPuIfPjPh9v5dWCF9y1wVAlaaBDhENhAnhtKqYd8UmkeIGe7kwhx0S0wj8KYRq/QIpqn34F/KRuZskKsYMC8WxpOV/xq0Es5DBvHxvbegEGF0qAsHyjyJ7WLEKHRQL0MollBuSC45Xw7L6X914dBPXVf6koSzIqYW4iMlrOxElwKK376uSghE1CNl2Su2Q+urMaOwtyDPpY7B8HZxkwtCLtx3mfyymKdhtPU4IAEwOrcrHRn4LFE2Ytye6scMxekdwq7iNRQdlcBoWTFQpp0rCcjknScDpcv+7ANFkbxRBKYJU+6UKx+5potUmxyepCrkTW1TUUN+txyQto7oJPL6iNgYnkEmbiB8kDyyGdcVI88ZpVEsEDBB7HnbbQIdpEVECz629S4tj6yJLNGK7xSN2BHNooNuM0XOK0J0yIoHiF+z2hZUEvKkgO457geyoJUKvtrXKQ5PuiUkg64YwEN6o7lsJdbuO+zF7HT3xPPR1x8fl04WPfspYr0NcPuJZSaJsMgE+a0MbOYD2C+fEPUFRAQ4GtT5BfJS5uwq78X5aF5AUW59Vptjrunfl+4mztcYAkaAKWkQ6EcG//U1bE0IYqudROsHc4/tVdGwaBGLhOYN0NseRyuBTOaWkmz1oK9+ZcjfKaHjKHWgT5UZGudNenur0fAK7GNX2SH2unv4czCgB+3Wkp1krcHPOg3ZuIH1lnIopb1VTZmC+z0cOMHup4eFaavVysD8ycamoIgdq+Lb/D0HOlRFRhRlHJwI/a+G8cozSeT+nDkRLvPto8xpFYzRMa4AuIs0pczL7YRjmdvQIVCPPiwq1vbte/3b/fMzBYdL4kRvT0R5yIkMNy4TQwyDHN3EQA5B5FwvjqH629U2PktDXFY8tCRJxAV3IBVKte5dOoZbbrIC4ffGo/K7mJCyZ9Vdnd5nu1jypTSkstSOl0SKr2EkrUGdIEuAN/q7vxclSkWM0jw4W6OWtnA2jevosM2TGik0fI4cUiFJDNRP6XABcIHDjtqbZULfiVU74wjZhXB4+HjuC5/6yEjmf+JtuCRZZBXpZWD/yKwZSDEES8elNIneCRk/3xLMSYJoftrgDizF7feHAGyVTMQyxNwPtnzAXEaUQ0UAfTtnF9WJ9Dpg2fwJs2KL2cZWEfa1oXcJcNITqB53WIHL8/bKp6s4Y3CEITL9oqR6hBXVgyEAjVH4edm78jaHB5I3bKq+bmOcc4thS4hRWaaUD4xxWKy00Q8GyCmKeFGiZboRAvi8gM+6jEpYIuhmuKWlbXnaQiBlQ3ohN3Cfnv3Li9DQMHrDrTMKt+SfQ/dKPv9if7lzIJMh6Q/+JbAOrUci9nYxV/3oAkOwsEjtizi0EkcA4jW8hWuZScuS/YqVleSGtShwcxtuxwm7zm/L8EG0qiOQz7SiAhDyUJXn3/CEnZngSA2IsbIWNxRzCBUp3Ia1HMNupwt+nmEBrBiJqaceKJ4lDexfFBD44vCVNGt89dbo9NeOsOjAohZsUvcNIM+9j3vnJRi07iNLNEpV5d5wOSC+sCv1Av602sPx9BshORGwDtnlO/N6JZwY6l2KiNmImD140ybOczP6hfynOTiGlMxQAb3o6aHDXISsmk7UL1HA69tIb2bllVttvRdftvHhGHyAp3INEkbMgyRvLc3+g2JqaDAJmvelJKGMppcoYjH3AhBywU4kCf9L8UBlua1GgdfduDfTyaLrW5FYuJJXJukMv4WZ9HOB3YPsHQPNJsQg4lKVnXZKNkGAX9vvc5nVXGOknLbc7NNi7OkbOab3kVVN4725Bo5lndRhQ/Oi+lrWCac05x44Mmq1Me2oOL0xnZgFZozMYxrhWSA/WipQZGbL48EUuDRiW3CSc28E8OlN2KDZEA6AKii2Wb/8HeqCPzY/rBJtx/e08usVIZBF5UDTrZa83Q6ej6I1Jsm9flWI1tpl2x0bWpcU3zRt1LZj6ldw3s88oXr2IfhrCBTFwPrCfeul9yJIeVLQ45RCXFDrx00GlfWo1sqncIM82SYFxkCW7C/pyaoSwhs65UrTWxYLcn8us88OPQL605Z+Cp15bf8kz7OH8mXZ6v1bVQXiVdCeZdfw/VwCGKR/8F2B2UvJm1WvUrCEZvriqKdkxvBF+JXqCAdYs/KAFUypiS2q/3OEgHn3/r6wi3GS/1zECVRTIzC+QOoehEzrcdTAFLJE9bLOLpPTO71yaKmIqvPUDao3yDWyoQwfLqJbfipvuciSGPFaYamNO3bGc8Y+O9BwVnHCHN2JUhxvtw8gbIRjB99BwXoFriIg6DctQTubEuUubkEjEDLnhH2pxlGRRcGFyfvz/zX/aqMzbvuwWTI1Wbwgg0WEKHNARba093AfTNfmRi6AV6ZnAR7TYcSye7HL0e0zhj6UBarAmdPE4qfYNsy2wP4RIDF+mgEnLabNJl7uY+kbuYap7RWWjuJdUpVNa+9Qbn7skj+60b4y4rXKhd8Jf7SG8CnzGnd0FO8fvpNokZ/0SdsQofujffunJCzVh7v6saIXz/ku98BQmn0Q7pryor5LirzcuFzwFLo7zzVsvYLATj5dyQ+9AiPEJf9hWDYlA9p3ZGQcwXGMR1ybd7Yr3A8TIx0RWjOd2fnw4Qphyw1VDd+4OHezbNlcE/k/HhAbCWEMWO5+xW7vUIibRTzkhKvsHRYYye+D2RQmyyeCnlFuJCCnaCMgdreDH+6UPIv+/AvxFySpIhxylBctT9Y7ekRJ+EWD/3EP9dhh3McGL+Mtd17AyJCoP29yhN+JxfT5gG4TRmf0w3+G/kXSuf5f3/vfY1GXGiBnpd28hdV8J2nw+klzFekaVck1W9NeZqo/w+qt4Eya6AfhZ+KdVW7tafK5FGo8l80oUpaB7HBgX8rMGz1j3W9gcTXTZI/Rjt+q6Y6DakWb7Js3mj9BASCkezGhvqG+cvaN8fZB1L8Mi+1Bd8xkHVKAEYELepuyYSsUFc4CjyY6sRDL1n5fDSmMdVH/l30cCdK/jlZmmEbgdpeFhViT8mmKQgWjj9xrkEBW1Jblp7Zn7pOd59CSVOZGS+zgU2ix72QNAp/CrAk9gN/bYiA9BB57u+I/+NVcLPfT94xsyXJbmf0GlTT5IRDpU+Lov0kRIb3E5WK5099qQrAfd4fTj9dOArwbXM2QreFXli9SrEQsI0MN/wCtjj2ft+TkJLVSbE2JjhKKhuMRjxFd+XxSwpDts4GMWp3ZMoefw4pif2X4WwsqoZTrfvbHq/XiSOPTOCQ+fsFybarnE8tgzbseBjgLziJd0+nPu8EWPOaD2YMHGobHI/V5fXkAvdFn/bY8/KdDHX/29QZkNJjnZuD33cDOzCNCxnS6NTFNaTNMqdQgAcI6EycFCPFD9HlaUtwIV5FsyV/s6AWBeM3MKGYaUSmbs/FKYmxT+RdItOQtZEGOk1+Ewotm0eVXNvhBc5tuOzJT2/TKtLkHyjmeUcdsRfNkopqQHGBOKIVzuNaPkwvTthLNrK1T20sMuThugmKdnAIfN3eTlMH8e/wrCSs5zTf7RXoqufCTLB8mUcxZGCQj+mS/J8jXvF2Yxg7dJX2oTAYR+c2OJhKi26l4PWSYCxMMUbQaFGlSViAYn7f4GhkquU0rY9nMoU1PFZ9AjCZgP1X+5DGg0CaKEvr24hwhiUDJf4tsQI6LMLCkT0//aPPpIe8yCMsV7sPuj63XedeAK0O/nmpIGY6/2+lcQlEuduYDP4Y6stFRiQTgUDdDQS6zopt0pMFf2VWWliDrFwzp8tRWQhePMKvUW4WTD6n+qIM687Mkrd4Qp0lvrzDFnYdk58CxdBmQ0kw2t/dv29CY1v6fDfQICDfFSfpTF22OOmEwcEzul4IifN/4gRlu2mZOxsUYrGIl3EOBliP/pUikWpKuA6kK+NL8y01RzcR8CIO2pr6C+9QICRpAdIoyjNnLn3FuRECJR1bL2c1KLCusLvXUvZ1nNAGnrFtl5kyfo+x1BX+5ATi4Fc8NtuisChExlZPgGxNjVF9WtgSQV8OqOHmEXnibAwRsrnigu2VnDYk3G4lwLM4npGECMZh8w7xylUeCbn4ArLuQLLBGc6kJv82Y/n4bOxeB8yAda7q6i6gPgCHgZAYRedUqrnPsRKcVVBj5jqWeFKEKfeDl/hbz2wd1jCLTBHVA2IYebR6ktf0Ls7dgqZvDJUZ/idqU65Yl548DOnie8ydP12QEVbFGVV3OEnpKCnm5qxi8704ggka4HgPHh7BCnmEkS1mV5/RyRo6VrRMtUTyIx1wPaRyGxadM6S1T1Sd+VnntXFhwftdQoylkA7jxpNI5u6vhDPBXqI5rj2f/rGYtCIKo7PivPeA+spijOu7AkPXtgfKKX/vB94rKbxmZeF4r+ve3X6UkzhvJy5w6BVviokZDWwSwQpH5i6brCxTUoOojxSwp9YzRLoM/CUdna9ryYJsUFodggX/ud7UUihItyhcTfg+gJuUKUbXxE4TrKYkcbA4cBJAwqCf0KvxQ8PccWGC81e+5O/NWgBLz/vTG2BI8VG8XgMl543D9jbYnTZ2fb/f+rOvj95ihAnQeGrGBW09rtnSCu/mJRybLgagzDSc+llLaW5MZjcjh0iSgUOom+o5O41LzV99HGc5MOd3K79wPNLM0RN6B55UXITsxYbN35NFtBM/Xmt7Zo5m0XfPLphUn4l/7/RqV5PReMRKftFyjMl3KajpcjCcyzEccDAiFggwnZawBfiWrqLNcH2j6Qdj5BIJOwpEjVnOI81ticNc3sidLBFPdc6KU9/AaF9uFfEaHIOu4kHvFrAzzVgOzBatCw3D3v3DUF4czDbCTts030tZpvtF5Hl9SiqpTcIb/O998be19oAKcgndWvgvS25grUsDQ04ICx8S4ZxHcf01xolrkPafftBx94gc2zh2kTTPzJaSpXkkankjhB7Zrh2Ir8VxS9/nhehBQfZ2F+rFAd/5dk1s6TyJQ4KAyreQYYk7zKCSMC9ZBI5CjmSdsf1YjqzRoMkQPmi6YYWF+Nnbqa+l6EL/U0d5X2zcvjfC9qQojoJvoNlCWaXnsi4fsEwzKMfw5C8hegO/f31Vf2xzscTKDuphGXjBMqkc1RDtZO4J7qkqcbzpvjRQwEn00zQsf2gKfKPU/SLyB/2Owa2vhNB1hrJkmiWPpNhwcDqm65mW9d9Rlpakp7bDQIhWwZJExWAsEyDxZg1IJqO55ItW/HkcEsdnRor7k6UUS3dQh46KYYmjhlu4XXbjtQiCWbasGuvrb8/gajlVIMfofTjoNeFCkXNt8CZqKGKdXR41JWWEQmp2axGUVy8AH1eWfnlB4RBGOA4w+a3124P0roVjYUSEakkHp6pD/WGcqh3FJ3xOyTC8EbzL20lAqLfNfp9nXl282WRQ5KHdSsYmDJmjFt5p5cU1Bgw1SSW9Zh2qwGaF3UkYcJ+a4e/VxdEVeEl2kuvdK/O4WuOu7VDGrSuMe++xZ3ggXhZ+RfsLTNb/P1iaSamUr5hCXpr6RZ0kloy1iPO1YIhYi70rzdceBLMMspGhPnMMnZZq/3rBGrN4Lisk8Ut4RUVIDCFo9hYY6bV2kmjclxsdIYzgUG3L3ZgyylAjk6GuQM4IF2141kyAvEAbTpcCd0Cwnglw+p9mSl71NXezOtU4e7jC393PWYqlqSheS9aHQEJVrHIeCnasCpManVAjSj1r+cXNEu0a2Crodkgk6zcjrbdYnpeqATSeJi2zn11jrKd/PtQr6x46LYxlUevt9ddOg3ds/B2uTjrkce9nN53K5xZl0KbbnbodyzckJ4uSeGsMdKQeLhFHYye3AE45TE69E0NSAbhFPWVvWh3mJsJBCD7zG5yakx/6kBdeeKKjGOqfW3o7LGA7E97PKh9PJsa3ve7PYrLipSf6Ad3pfLYw3ZzFx9KsFWvarWxnvVEt0Xs2is4ag+Ir5QgU07+oRRr+rfB3OnVzZU5bAmNMIp7YkR660GrQtCIoB8vt7V6SZTz2ZvgV3OljmichXd26r9SZGqH9sHzI4cEmcBChHhXHcrGrIeCQM7inonCSCUH5wTh1JGX56tQTNLm525XojKpx9nQedaedXjAbaqsIhxJhM+Mm40sISmX4qvRJOOKbBIUup6U/3T1l2xVJTPM9ah4lah4jSVseF07dq0p/9933epRzBPP4qbKAkGqvtEr0gx4464LJWilGt5dTDvIR8oFeDTqQwG/hrvE2ks7U2T0UmaKyV+vb+iCPvz9DiSpVLCjLf9iBG2ukk93GnDG3gqHMvPL6Ew1EIZ7a9dic64VTc6+S1UgRWiyqnxh2rTUhDFMHppiNqJZiGtAASuufX2hFshCs9Hxo8s2ccq5Jz6weyHhTSHCRJV+0X301dY+1Jjl1dsyC3BfaW1SI8l0QS4Av3ozqDynqO0weXyv0heCp2SQxiKNbTA8I4TAtuQ+poS4SZ0s/IB7djE7du4ptCh8XpkXllpY2X14iZ7chzbTBzRcnGs4n6oU5/foG+1uO7NdqKmptszJbiwMHO97cg/rnzrCcJs3nCOd8Cgbn8aGthlZtnPnQMcpmx+iTf3BnS4RwLY5qd0MdRwwMJN3jGP7gkdMOCtYCEMjXT+9e5rwOiXq+qoIEiAJiUnYu3puQI0aH8AO41ZtpOONPMMxDp6AyW2KqCUlcVuENCMHlohn2bBM2oy1FWTC+y1kmdJRT/WvyuyT6pBG6pdy7clZaCVRxKBBSYcn7R6GTs4ApOkcWJ3VQG+Wwu7olciOntu6U6zVCCz2FRVfjj/NrZHQ15v8kQf/JhAVsPwobzjXlDfZpW1PEMaYwBPgFDOfISY4nlYQOfac72TQn0gdIoDxr6MsLLlkm9zrdvFP8dlYxy9tlukMyafO/Bak+Y11EUEEnbNPxmq5dVmntev3p7YohOxHmp6zqpdlgOknzZKc2v4nM8WDjavFIqLjc1uJA2JV/mxSF1pSo01+KZm+CwpIXUTL4gJsWzxfXbHAqPfI2uD1DbxRbQxa+osRKw6jVOFs/UQSN+5FXaaIfBhpuY/UAZid3TQZogfekGsCIiEDBroS+PQGsyww6psX0bjpwk47Qz2HOLq2ddaF+IimM8mboiFxR4bvYRy8zusITKgYEK2ss5yx1ShoBNafueJhBvglU5A2EtDwUYp/5XHf6Vopz6BCX9qBWOP0OrRljFxtuh1OyDWqPO9x3I41L9z4D7Oncww5WFX/CwJJde5hD/B30NgRRpn2SLC8rjdz3QWqy8S+oHhzKJ4ZCEnBZgLkLhG0rok9fUQLF9N5s2Wv8oMQ6ZoW7JYqlXY+TbDfHANydH331jbMfFBjw2XZekGbbCH8JsbNDmK8qqQyu/zD+L7xW2W5G+XtzKj4pq+sEfaiIrOG8gX+eIVWPt22WhXDc+hpg5G2i030NOCy+L9jGzV4VjtZDPWPBRXRiqIEnRLlqJWZMr9apnHRG0UCEhQIJ4zOJJf1GWc77FU3am8LtJZf4CpT6eKZmuKzT2/ShS0bgSy/9zkZTZ5752o77izry+HUxCXKB69183nzBuwHYztV+5Bo8TJd7bN9ugKvt3csFc3g5kuCXKSt3XGv9e4ggdokH98Q1FbMi3EYpnAdF7vGt7qR66QWA4ypnX3WpZuzmtZzpYxmxgth0Wz93cCXq6zkdYLk4HC1VBmQLO/IVMQ7HXb2fGiRKpbXAHSvgUlGwr+P9rRSDN42HHQpn7M0slLCzzpydVvoaO3ZFNFmuGsJJi9IglPXqLa8BKWr9yVxyfzn3nf0b9AgYPYrBXR/eY50WSKyvx3ceoPTSUp+DsgwXBbpTGOz0FpNn0pvmiL+o3Wjm02AXWweVU4tQc/jBUY+6eDt+QwY6PiAqL7qtEuBF5sXo2oLOPUbcaNmWDI+rwyazOvrTou4JN5ai1f6sy23KJEenhRmbEeJaBXrW/IGzS2QUbmY/biFhV6P67buVjcKs3h3lC7wyb0q/kOOXA58++fbqk4+rnF6eY6DmjXqrlNXtxi0Vfx3QKYa5B0+p1O2V3gTHbaIQJxrMILEceCykrtAaFbenU+BDoU4DKUr4xFOlxa9FtkCZ1itKdBQfmYMTk4qQRHS0oDVwNIHZrVXDVD7AIdbQPfW3dfwg20alIPI8BiXZmQ4VJP7QjMroUcAQ2McUkVHzL0GQ/0TKMLyYsMIHSjpobUodKxvZ+sR/x631YkNytigPydPRIJi/fchAhl9e2BSdk17Rs2zEpNuxCLih4y3wE8vt/gl9IMis969euZgI7XdJCAH113ARu9Rm4IKkJW9gobA9rEIEVdRJY8ZgxsMWTwZQmZnqXDWGNrSXM0j+v2F7TN8k3d/DIsP2qh71u8/si5dryzpLwGmyAdRuUnlwDxKE6cFNKQcy0TkSiG5523pJAR2lPYmceD/mRFrHRSERpIsoPyvdVUJa1fxal6APLcbvCpfiFbUcVbMyh/YNSKZD7BB2V/SQEjOgB4JrTm5L87Rc4DPD+lLj3c8PGRQ+D1jQfpWs7l7qnFaKiWZa3loeJSQFKY2EdLj3C1b39I9jrz7TBsgoNY5gdlvwEHT6WQrwpVkRysjRy1rB9G1oCWWqr+TQr3o/+BRYgULcPIuEkRQaEKTkLJR44kJKLwXxvG8s71FKtm1XKpA9898tsIp/Oy+kwjN2rkcA8VxPhG+6r/himuNJAJyLAak64jDiloU1RuMj/llo1c+rJAZp98/rD5UzGcPig7rCyk/4yLAE21oVYi8Z/YYSOckFBhJhrKAoVPQX9Ho5XmqQzVqIJ6KW3IfZ7MPvEk4t14P6LVn/mm0yLFquJfPr00QCS63NdbK4WnwvieX49J5vWil5iGVVxqcKFVe8+Iclxf+3E8QYq/PIN6zkwGCD6g9lxIjbPFt84mJ+x2Fv8f4mzCHZijy0XT/eMXvzotYEt10rdKZVUi4IaLLuZn114G7dpQyg8z3cjqLG7s50xUcE20gI9GmfBkt8VmB7xEn11xdlVjywTRH86pwCfxb5dW2laqDM//zrn0JhP5hI7ISuL04P/PhrnDx3pBaK3hBJdBeElKoMF5wk0TXtlZ6uh1grbVQ7bfXBwUHMbm6N301oTX4ibR90E31hSOh3djnhS4y7eUk/LPwbjdr78pa6/Vaocsa4xb+FzW28V81yfizFy7/DqrPYJPe8pqqlAcGrYeerBLLG4Zex965qKYPyIqffm5f3I0v08g9l1FbhhXvV1Idf8sQkhtsJAPfOlvHBAmG1DOJRUQmAv9y3qwWohkvv69ZShjtYJleqXDXcBt681hS0YayluMiQgJ7PHDDK7erp+TJtLjuMNf5mrwfcTNBiklI43l43uIilkEMomOF1j4axguDAawWcyuW5JaMA4lnfo9JvbXA/Pf50TSUB6rmyFTLzc8wGMVChCfEwsvEzo+PE59PyNWr0lCwFpUOE2BuuWzzH7UbzcoL3raquZQzJdutwArVeu/12NNx1lj7iwp1abCF8MnUs44eagI3OqKSevRcl//Pmwr6VZTjWPLE70GkRuhvXPOA4kGNgU4LqprW/achstbhND6ooPNM129W/K9IOmgHXQEhOVHbFxtd4vqv+ZW3BCHfJT26/YNZuB0Kgw32Rc72aHPX+WNogR3LtSVx7HL0378qmcSQNAlvJHIMx63buTm6Dui7PSqew2Tvo8q7MwhBab0Nenw4e+iu+DkqYKG9Qj7MkHKCDgt4qHpD6smIX1lSTzrw07inYbHvWBuQqtmb8kX7qo8wCsSCXyzBUN8G3pOrHFY2P5VubUfdgh2dib10YEj8ipxFDe8HIu7sty/Ou5tFcM8h1toVh3eshlrABvIjojJXOnEQ+EWPnLTNYEbmxV2iCLD6RM/oytaQ/66doziBXGHZyUG/t5MZ/jkwa+qZKGS0j6o4GJZIuvZGfydvUkH7XP32LJFenyaeQwwFmRlbUHBqIDejZyFmrm7B/UucVFScjKSrDFlpSQOAJIfnjedQmi0VJsUmOXL65Q+jT7SZTuy9P9Z/YCYjtcKq1hBd5CaLPalOsYyLZzLFzXNsHIHcRejVNadRPmyQg42tH0ZoKXtfLiD+eDg7JVSUfdjyDgaP4bMzkPud1oUgkvUDAzMAuufPzrZTsWJWJYmCUjyqFEOwa7qrSMxuSPhIbyerDp0KxAKojJ0nMKRJ+c3QXc88gvMiGLIxlYVkm+WVsQ2ofZ/rgMWVt1aq7K6EPw0ox/Jv8/hnS7CwO7BVXTTBzYp0ui9aYJng9wtgV6togobXFuKUgHbwXaddHbzn6+LBFcJDHAcsusrZ1dfMchrox7sC65y7reOVTDv6jqruF/2uaoz2AWNu9CeE3Hjr2TcWc54bmUcYVo75cmX0qTim0VF1KjT8Jtc1DSpCEdIDvoZu8lvjBxaiCKejhHgReaDly7sZ8yGYX8JWhxSS5CqNZJ4NR5qJjUEJzi85y/bqm+qBkBSZju/UIlaeqKthqhV0FDyCPf/LR2SUlHFxBrwqy6YmpgIQl7dVw6lCaze0asUfm4xFpQknO8ibHRcoSxeTyxLoS6AX6t1JVvgcBSPciVO9Lk3aGZk5dCB9ODHez8JBFjiz6y9EvQBEeKEzF0ax1LtTm8T2Vdp+WlHzCdbYvWVT4gQ418x/MDn4r6fcf85nHspdI2mFputBoKWRjymyqESShFhktDPY+nLwPoBZTdM2k+O5HMTfHywdjl1TQrjAO1fZ6ozg84dlUbc4zT60mDC+DyxZoGlLAFWO6nHP4wBQyMlJEFQfDtVHaW2+983R5NPYS4H8nKVPCYVqjABjPgsnfkygehU/JyJzkmYC3wm21cgh3ksWre9xrL7RplC9sUaTzLety0VcrFnIBF5za+MRpzEcPhDLO2BlSksfvhO+RjlZCxk+wI4ktNoImoOesE9rtCMNx6srKbVyOQAjgVSX9fvuF4vLrHxR5/kbFviGWoM28iw7TZ6ZXFJfUhg7IUOw1K9/6fxakUqV0lImQnYHiXOPXZ55iTKNDIHN/A7XX6nSm2n/Ds1eQKaYerODB2n4E2B+y931USHD3dHFtu25nrJH50O5hc8/VjA/9bWkI3iaJqTHHQrlcFshUc3+8urF+fThpcMGImFLI6nh1Sb7XuBrzmTMw6I38KOb8MShFErn9JSGo9rofano7iMRMrl4up4RILQbS7OZfhbBeR/qw4jHHkTskpgKT6L6a+Fre+45ZyKbA244ChFBo0cWsWjLaFFui0BBDSC7K9JvHgy35+fOClBsbxEo3Evn8kM/IsE1UQfC3kT2DvypPpLsM0i88ETRDmKnoMcIlzd1gcgOSxP3dP/3NLubceExiqjYtXL3O/gguTgyf9PGGq+Raeur1H4YwvrAPsy+98C3NNcECmMNRsda8phniYujQRjoLYfRYyQ97X4DULy14cleaSfATfPv8V94llWnBLqTGFx9kznxc1Ru8ssU3y+cEdr4eDu442kVSGi5PQ+pZ+HLtLvGQqJmEqjvgDB0cgiSmbEi4u3yRhRL0Bv7kdZdNhSY4M0UW6A+ojUHGe8YiOGCPPOpZKtcfx0mj+zdjBsYp2dn5rKwc6aKF5Y8mfNF9RGcP8LmWDJIWGlkQ7DvUDJdA1aqtglY1WnbCWrg5HUcDmo5xiifbWfwKYBXL9SiGTZgkCFxoO1Fs8PwpvEF9LDO4gnj6o8eWa/gA17lGB2BYkvhB1xJTAGHzJxt65mHiVZvlRnn5Tu6QDzZZ6pOuAA54T2T1TbIk5I2zOdY+5iZUUgvUTD+LeVbIJ2vwpK9ALf33UMB87Rgn68jviAPP86WxqO7G4vqkPR/r6vOCIR/zIfxfW0SBxsiwYaCsjSHG2ICVBITWDuBZcaljVX1NijoN9OcREiZX2agdbN2kd2ZUQtgamPrDay5aZuC00IdNCgNG9yMvVZEmITQhpr60M4MYJxzzueISV4CknnYsNsBdcYCcU9CaoPsp6WZDdZAX2u5euN81S0uBjpxdGMqKLwePg5BcJVohPOihRootAb0Dj4XMloN136NYtgCCZGv4UHs0OimKm0vUGDu/vNt32o5wxxbQ7w/v5KUlwvSVRKwcpK/w3vdArXbvI3e+lLQixIZ+jqoP7lVJI/hiet7bOs4OzdPgwmlR25gI4ruS1Yos3B1iHHFgSFjtE6yzO7DE2KdBkqi+9/EgCKgVn+Wa39XVgavfZPBBjX0DyjTt3Bs2V5XYhcB6n9/tJg1wJXoDpmdsVjIB+KCmwseNuYmuhC5oFuIO9A7xkt4pzQV4OAB3D3ubZCyfLpUr20O6luH0iSeN6yeGu6Xs91J+vWufs7hr4NWoqMVvkSkW+y0ZjSGuxrgnimf/0jeeC/LGaqjhopCYZj5rgCV3D+QR1QwqkHqpJHxxfymGh50bbLn1wP5UW6MMsObMBjgZsoLsgacEcHZS9myf2FkBesdL/AlWtvkswH9XsyEF7Ne+7XXV4qNljPm2MR0pTLjEHxg8zPqw4HWOVsEwu5HjglJDdnL1+y/tuFVLkyTpJNIzIRXr3I9OUbt8S3v3kSZv9eiU1yojkNbU9H7EscDVn8M7syRxkC419oCZFHLb8wTlZtxoRTSE80cPlvuTyI/EW/6MhKRYYkXFIQuYzbFlGCPHJiZIzzYGt68XiYWWrdPNI+TUCJsKE0a9QFw89+rTUPdXQEKn/vuK2ZxDVQd2e5jTr3HtMaG681XgWRr/LpVsXLebW6dKMCwz3ZKEPQ+ua7j9z59GX0YRW/bBasMDIpMjY5/k6hiiWteP/vBB+XBWn97H6Q76yF0EZUNJ0/5Pi08lD8Ag5A03GSKUVvGyByagHK1z2iuRvx3pveCvPXMA14XOYx2ja5GgCsqSrFPl/4dPIDASU/Auy+/P8ggfqDTcV+pyfe/K0DqMxQBRgsZa5sHiVyEUbj5cYEZl15FdkeajiOUtNoudfIitHLBI8zvaULE2+rjvXDwRTPMB+f+Oajj2BXbTg6orYj26fIsGUJESDPosfAuH/FDc309eMXdyz2VCN/QIi+M0TRpo2OnDAMjlKgNIXh/IdUl+3Kcl3fDa2QbZOUYl4z994tWpbqMoidT5qUommo6+ci+0sHRi6Yglu9R0LED/34sCg9SMxYoP7vHrxdfv/+RBeU1lo7XxX7QyARSnONupyBdDK32afpgk8PeoROpLYlNgTrweaHYbcq9KEC0HFoCZZK1KTC7leaLkLBzcCRIZlPAtruw13sLjsnh8ZNGnYD8kf4COAJg7zPnWWG6KiN5kxls7/gYi4Mmqiw53eg/ucVI+2nEL3zhGYorEy3NgMr/OavmQOCETWLIRFnU5E9CcBdwqzvJnWWXiGr4AKI0y15jBHal/n2f36QQ8tTczF3QbrX1udiM2Fa54O/wVNsT9qTB3bGxvTR4H8ZPXh4s1V21ehr4AnyBS+InNLXkzpXaDaa2GQwKZvvkPHJUk8rEfh80oLLxPGY1WoYD2rzFWCQeNqgUcDm/urXCTGsnOgqXjTDrEtQxjTKejD4wwhdJy/yvMgrWNfkePUDjcgBDbDTfFA6zj4c0GTk1JDBuJPuqT7HHKHPHCK+0Xp5RHr9DqBYO5nkUjeP6mqpsaA2s+cSnATxwYLZLNqRgu29I1YraQdLawy3XtMDHYwU1ytBRAjZr+arCH8osKp33f3p4HBhtRWuhJXjCY3lPdouvCXYniCnC2fAJFK7w9FHfZGXPkmNESNuc61rwZDIqBA/IVjcwACRTli/rKjIGcAon+mrU2dogQ9Yvk7y9SZIzI48c3rn698icNKiklO5WIb2Br2u4+SzPIojsj4ryz6A7bu6N4LF+bR2dyk3ZhZ69Lk4q1HfIoYYuAHZTjVuTedQJGrz1aXH93REgrO7W0WlDIL9/ApXygGadQOlNROO9cYZAgczucJdUVb6Iyw1eNZtPfQHrx+l8C+CJR+6e1CdBqDrh40/KdSwaboCFKJ6AgduF95Xfa6T14ccywtItwmU121JXVGxjdY4QfAF2TJIENppLmoR46yR+SFY0IlgrfVb2TP+kDyo4r7wVTV7UhlNZlxTwctxfVLE8gKDFcvIWl9UP0ByugYdl/ilmkykksSLXL6B5Y9QDQYte7/n/q5SKWkzFdwBxyicrOM8lkiuQfmD459Rix22aFoQB2uXinJ+Kh0fdKIdVG8EuiBHguRXkuKfgapY9YubPh8FKp+LNNqImBBnBe+Hk8p09aXN55A782jSzEVsj1v6ibgebVApo7hkpWzhKQbXgLSSl3iAFd8t0E27Ob4xxAhY1+cbsSk4gpMqZWSTjEyy/AvOJAP09FoqBPchjHqYeTtv9dU3QVIOnRL0lG/unV7xwGlSmrC99gbsSZPNenHFDQYQcxOVP43wKYbYKOKn3xmEEaUZza+3W80myvFKjFdbUfVUPzAy02LUgQE1zIihCKeTGDRNXN/xBNlE/V62SkGbL41mKlIW4KmUdtySvYZdefzmIjtRTo844ZWTvTkT4eaYEizPLD6pNI5TL43+TjgNC2I8hKfIIdeCqvF35U0i4Q/wtO+I0+IV7R4Ffz3a3hFnP6BSsVjjDCyBk97uTNIUdCeEQ+80AlwVjeLXeJQ9lty1Mop5s+xeY8XKi1zkMsml065OenyeNE+za7tpxFYnjHlcjIm7FwxgzhM/k1LPG3lWJrwwnqziBJjM9/Sk/rwE5CVJCca2FmkXQRDUS7OCSv+LFDu3/Ux69rJ6ZhTH9Lij5cvRpT2zmunMCgTEGOj6C+MPLo3VMsNq/BUy3fhL/U4ezqTMQ2yPjqacperBQKV+REUSuoMABk4zXT201YjyZmAExzJTWwDCWjQvWSEFqeiiszX2ataUGuavOwVr1Qcfi7c8Rs03cfyLx16cmpdinPwOxi+BJq9t1WMzth+z/izW0NuO5+MmMKbdWoKUeVjmtY7Y9twGauHfCP9V7o5fTafMyCHErpdur6fEyNFMxdvIp5YJvivGMw7PhTAj1QvNMOWNsmpiEFNS8zLfE/h0fEj9zQ9v5PAPyflv0WNkE/ltUDAD0Tt6Kzk5TP6JBDQ9B+G3sK39883brIIm5Moj5zSrilQI2CC05DHFQ1/7AgL0N6omz7Xg8KlHLnklH/YoFsIW3cE6QbqihAXe4NXfo5KblgCNssDaV6tNbwW9vrqT1m0EyPlnuInplfMAeGpAVQf/kR3wBK7cN4nCMWFLdOcLK78P/JQh4KfsMGLdCompJo5LyRzRLqxtviJwtKgJXf82IeqWjFlBQBLtI/qjP2IkC0v2lupOeFim7TN1fB82z+E7qHueNLgH/z3eJ5WwjKA6FSdRFY96C/2F0dC7wKHKJnFwxGJCuCMq9qKYbEHzucYLdPSgr3TUfZ6Cv1bzrWs8ysL1PPCY05hR5EbW3w+DRh06NtWrEidjbha3P0FFS7f0+nXMyaCbAJjgmXxVbgvQwVXexhO5paoUNgkh8RY9S/knJ8d5mwEpgicpDJMmLVy6vECMCTsc8uvJEWhe3BYXrkcCqL52B/6neiE5koaLF/Ud3bEtPaIDQyGLTIjU6w6qgSdt1+HOndM5ZeiK2cFEQB0Y7RFjwsVsnANA3jJZIB6bxMXZoyVxZX9nNIr3PgLGkFJjANTeBc2urUxT7tF6vQDQIny2NBBvXdbBBHnykMbfmlvgZaVV2WupyKj2SpZvDD4s2BfeVtK0cIr2OHo5mG+HpWkvs3vfwuf31EPj1R7wVtCj8VAFKuqWSJ2t/Dzyu1lIDupD12SwMgw0TyoJgW82kjHoePM/Fzq4IpsZmhiNVQUeNudRRGeImByWAaEmg4mCOlnFwRz6QiyHzRBuhfJ+YXchVgbvp+AO4Mtzkl8vYrDz1yingJqUP7J5uP82M/YlFlcrtBRN/rp/l5AYw97Rk8a6RSOJ3m9OySz5nnmbO0SQ58j8Mkffgxx4EDRZma0ptRfhfqa4yWPdBQNXSmkpVEzJR/Zeuo3GxWHdGZ/qbdQEVJt16cICHyOqDKzxx762D868DzRC1RGltuW+mxIWkl3Fy9sv+FbyBmxNIilbVon3e6YW+z6l/v+7ngWH8bAKOo2p/e6fzI/jZr5tKVIiT/ahL238bZVpjGqmO0Tt2h44w770cxWqlbpg/PYWVH4VLSHw6PfyknREowL8nrEAMwCtVBiBX/6VOdVeiSZY4kcaQoE7XKHPRfU1eYOgF1jLjwrE/F0f42Z4vmdFYVk7DAeDqVomjSbYqGshALFwyoJC+vxT21AHXKr3+rO/U84L0/1ovVudeIYeK7IUzpbJ+yyWjKHlL+nUVVH/WQ9N5V6nf3gOahoHE1PGHigmzXPp0/CgMPrKl5NhZuIfyKYi4qVzcqUyXo57xIxohHfytkTZO+yf3sxz+czybMiFAh06r/2VdAo1iqZJrXkW/KMO2lxaYmBDjwB6kTkn3G4t8selDRBLDxyAiUQt5v3pdAbdVqtDjNY3jak7HvFocX65tIt2ssptVgRMOuU/pbMz5C3GH2W8BOyzr9b7OT+BWFBF/5PskjWI6VWAtPV1CiKn40IAhX+JEQDMlfGFh6eME+3JVQszq4PScm418VZJMtHbnR2HE9tM4KetxCQCxtp5sY+APFrWHJP7mx+OqgXHp7vJTGfxDL8b1u0rcPpbPymbNmJ/iZPhysLP2wK5s1hbXQZ2TV3YNyhdNpRGuZDXoUVTqhx2F47LoVk27qk6MQQuKjOXTmNibWGbI2OggstYAx7kwkjW6Ub4cIH/9UVlCqo7BN/NvNh/M5u9KpOxnDQB6LWXzXsNLzTA6IemF21cfnMljjjQqDF1qDzZ8C1S5tlEEZkOmwGClhBzj2THpj5sP4/t7YjxIuUZunB4Uec5GbHegiuXD7g6iOACUGxOM6sioOAOCVfNlGqHi6V4I6esFL2RERM8tWaU6kc1U7aWwV6Ruh0tQysbVI3SwKqaotNSJRZMt6QLYffF6ZRuG821U3EjzPddp/AkCT5zdg2ULCL7AeR75VXViDK0oYwA7LSayunu17PRSAqTMomfsVIqiuBvGbjZW2wbd5mLTU6QL02pSYv2Mr6GjXO5eZ0k5ujvKGf5URvONAxaI4HKeK1aL4BxdTY6/GLWjeIM42+h6I1hqLTSIJlD1UGE5lUbIRsrSl/N+3O020mVrX8r3Ch1nFJO2D9xQPyxYx96n9xfb0dfgic6lMqQTgnDkTp6NBbGFWOIV54esMIxVSdxNhdILLx5ulbq4yBY/8onbb+wbdMpD/krjwL5rgF2FSkJjmpPPrRdn5hpq0CLCXCBwa3sNINT8ugJcCd7E/VAfUBXKkemJaRN9Ge3oypt7RvRl9RHKFKQXQCGSnsUmBTgQdBEslFs4LFYqipeI/Sam0i6jaatD5V2pdgI13KkWO3XV/JhPunYpRu9zPA+LVpKdlTVGGyfLsUeWms2y1naEIlTCZuJNuMpB0c3MAEEgxXJ7hAevqdFwh5oBq0VZ0/OJDU1gBslgu39TwB4IZaAj6NO+U2Ae1rUV1pOJWDqoOtcysZSAF9gHExlIP4QDFGZZCioFekR9EKNNW33jBHMzPxxu6dR52fxE5wKxrtbg9WqOIpFBWLLEJ1hw4+4M7D89332IZYeRO31utaJ6XNnGI+ZitI7g/PXtNCF0ZYbGbBb9pB53trwPo2qKOd4lRdkCTe6hpYGr5QqWN+VWluodJt9LwAbShW9sk15WGwusrXLKCpQpCwzzgfeQuDOfNFpjnQIiVAucJM+oZaBdnTr+t4V9IZQ6QDwfrf10IfhalBPl3Exsy16ZbKVuh8Sy+/jeyCfJ7QLGAGzQMLywtnlaRt9oNF1PA5GgBXnGvFtQXG567nHVV7PwiiexQ0G/54DsUsLnPiPl46dD3qc6WU3v8iKjFUELSmsQ+QAyu58HKo3pYWKOyHJ7ci5rzf4Lv81flnuopnIxWjOTNUqofIkYjGxTyK59zNZV5d4ndGZ8t/6uW1vChLtvUlAb8LeFPVoCl/YvMoYYqDrLFGthRkpqC8GjilAfulTQ83YPVqXcv4B4+a2blBczGbFmp4J20bEXVCAw7SsKqXc0Vm/2wKVWsn6aRvHpPpoGG1m5XZpwVx3e75W8BU7jIFgNMj1EplskOLfgQIUMQAZizQhIZJkoH90JOFPeIM8SuNHATwT4pANskcljuEEE7DPAW2zUHoE1SpWH1f9kbbONxnE4fB94r9WuCFyiIUMNiw/re2YSa7uyyzla2SV9uFd3Ls3e/GpZUO4YWXu528au+0mws0v+5YpKT/SjqPzBh1Z0U3e6O5ESj7uliACev8ST7NAzrnISUodKYpMER5kXMuArfz1jonr+iGd/FOyau0qAC57U3wzqBYIZE2xNZrgKvQX/r628JgHRP/0xpbYkRNMiiI+uqpFfo60ymxi7BmWhq5WTNbAfABQrVXqCyejye+z02U7MDT0VMmzGJrhzwDp41ptY9L41INVkD00XTtAl6fKSErH9oA6o3+hQJFab6QBiJYb6+QTRu/HnQPJXEP6wCAeJ8UdkHKeCr9wdjCQpRA3GhRhNGRVQdxwRhtFQH3vcu0jcovsf08ctkhvCoPjbUog7lH5dDMoEPhXf9TqIpEnfLuf1VFswsS98/Yj8qVvuHMhGrQaKx+AEmVEnBqDcN6FYtlZtAiE9avLQYryRtdIlr2rHMf5oJ43mme8QhjO7Y3Obw/hPg5eMJEjWTIdRXgeXE071N0VCMjGdJDYNkCtH4tQ7WDIDHpjj7BtIYZUvTGB0eaaW9iw9SFtALhFEPobXPpG2aLYg9W7+B4nITnSkD29d75aOKTxmibmCpS9a1pOfQCCmz5ikz52OKZAeW5Atw8+N11q3AzpGDLgxhbkcXZLt8Gn1IV2FPwC9mykGz8AtgGEKdMpRLBy6PeWGymEufzc7OyN8+TgK5kNR6UnRtrsDS9hwW+tT0lqjwF4zI1cDGCg7BvYONAW8qf+LMKj4VfHSf+TJN0URwgQKyf1P3Eabry7p9Mh13oj2LielSZZ8mMwxH6VW85S7sDVbbDqBDDAwbRdbTVdBdnATuIns5TXXCdSZurr0EQs1uT8gBEIYA7/qGzOZF9nqDTyTIRFT2eyc6cb4IYcex5sY5cczakrwjGLS+CG7THX6GH7eiXAavdoMMn0umH61sCpPNiwayFA7UYP5LtBeENKJbvlKhOxKjZiz+F2mn6bPSQkD9DRokA3/m/HR0iXpeim5+9FyXD1T6lcyhtWWkRWcrYH5/n+sExZhDgdlZCqUCQshJkOVzEE3egVfEwuIK70BAFUfa1tbgwXNoXqzFL171fkxjdDDAbxMEQqO2jvJZG3YqJjNTx0n+JIpyOlzBsMl8dKPgxN84XP/VKEdeMQEyilw2nb8HWhZ1FLnbY4+ZIAsSZXVFTr3ePcjw+yfVpIUIZj6wsFI6CULS9UDiHdtx5D8+kL/sZ5uNCmFCQthl/AVtPs3KEerIcEMYL8pAkxN1RB8qVMJfjWxkb7C04DTUAaUlwVEHxg6LnSBVdsi5nAeNH9GRKzgwhBOqujCJsjNsSlrDYYUB7bBwDY73NMPt1XzaiwrL6SpmfHln2CB28xtK3ISUGHpKYKV5fNaoujfuMXmQifCst5oeToXtMoBTkw+RLRCVba+ATWA78PGVnw/gIdTyLgKteMEQMkUWl/dbXjxG9lt6QFYhgfkBO2By4zeSJJ3r7/p5MTIQJoDW+0VmewfWVAktzvzHvJUGghGmftewiPqGFVLPtAmvynI7pJ+QCl0rV1J15qK+8SX4yyh/koqqlDPNiteC29ZMelYX0yB15lzo2ti85yy2qsq++LNvXujVe0rwzyeEjXS58EgXaazY1f9zb9e2+YPMAiEqZIgGNHay4ZLY0NbDviANhTuZXTyUDFoL99upBMpIlqpuZ6F3HeMEiqWC6sdkc55Tgv5a2zVVTGgzrv+wE9ZJhVTpZn8HHfMH1BlC5yKyAhZzZSouZIDzr6r/RUcCLeDTzuU3gXXUWYUpEzOtI0mcFyAr8X5ita45ot8QMsL30nxEhMRlZi+BPFvnvnhLKlXyFFqaTWqgXeffle+yRfia29CHr64nfEJC4hT/YBFIY87UkkgRQn+QHbwFc3Rv+zIfvdeb9ID7Y1C9ZeSe+fni+gT63xzvS4taBzuARAdU/gxXHDCGr9Lsa1doosVsmTP9ToOJrb6bYSaoLssZe0yhx7ORqIzblcBZgJqjVygZ5DJGD/5NCzUPfvCF6qyWe4njNrKQjevlBJJw6ORuQ2fDccC0NKoNa5th3VsIwmnDnmUvekxKA6bJ7BocnzE4NNSGSOUYs9Wr6k0HsMMwlA+kx0YGODonGkzskcmFFIFjY1Io6pT2KDgQOZ2+k3D4BzxOSYc0GRLRc9YLKibvXGFQHHhPiEfG7UBIz7GLcsePl9r1zGe8AY/JGDQe0J9QvX+AfYuyZPPbo6Ytz/eBbsbPbdS59gVfb0nZ9Z/xuogF8ERn4nmGnj/KSQ4SOknjQFGv5cwEgXvmTd/qEpDXHxwMx2ivnEQOCliQfhYeuXaF3Bp6OH2nQ/uMTtg5Ab4ojGRKZ/3hdqrVfcfWVAufh1rAJ93DHzLh0A8gbhTTVDYsNNR6obVNjzZFnAmKy0EB5J2IdO99MG/J9dpnV5onRQCFHyCCc/5ufLWnLjJeALHQuM+d1zKT3k1AN/I+BHbFMCciv1144SJH0jm1BkaUhaBv7xclzUdC5bdWUhh6g8E3Wy/ABk/kFPtszBE8O2g8fGVOhkp1XaYUe/ZNtxlkmVpY7UWymqBT4dgbsPL0vhDVD5h/svOGtE8S0W4+O+LsM5cGdPk+X+BTeLRXDVF+6rFJsgyWd+W8zz8ex1H2SGzw2NucKaS/r9d+w6+uEYqqc+ae1vJtOznSl6se+VuGLxShb3TPV2jQLa0izwKZOOjtiSMZM2unTqZmtdMwVW4owVgPAjqyL4A1zqxqrQ7Y7HikDQJoOc91DN7zOh7YNbPVx8JlrCQRCtQjuCZJXww+mV/AWrhp1F7TKxWIZaxy9jlLNOTkFpXpu+177wsuhkMSakVmCNsWED6Tf1bK01uQkrO4puiK1LfFkqFBIXIoLlA5VsrqGqWmIplqdQKSi4h3bpRvwr0xZ9m2oeqg0fymyAXsjW71+UK9g4o476dpLxOpCvCZDBO5NBqZsKfkW68KQhr1h14Ea5z3KhEmdu3XMsjyI3iBbHat2QFIy52ZGnb7eGHb9gmU8iq2RgRWiAXmjJb6mafm+48p78ZZEp9+FBXwiK4gpcennq1lYJ6zd9qLfowzxia+pN/dZjhFaiUf0ekb9ko5jzUiHPL4kDtUf4kfSxBDTF7ITFQoeYEgsfVZ5Jg3wOxC/KNDa52/S7homHOlB6BvBX5vnCRF+eW8ojL+dlNd4FNfl7uv2ZDbspxJiA7olc3iqz/cEJQaZ2na919ch62zK8BfN06MXgXJ7Ads99aDpzRWITSuEybSt0WFCxJndutz/udm2y2MSDQCnC0mL9pO5R5nXNWksi4AbIuXvLA56uZkzOw0h2ztCIGz5josOrvDLKJOAqpnjYuK4xBPjjLXIW0LYsfLyO7j59EzEmdF0YuI49qhkTeIaZ3zThC8DCZFy16ozTdJMch1Yjp8+RZJtmvnSSyJMwuujGbVg80HmaZCxDwYeSMnl6t+98OlYbVd3tJZDPuwNtunZAHJd6UmvHfSt4hccx/AdlnBepNC/6wyi07Z5SWiYyJU9GO/yqpD0iZG3kqQYNdmd2WVPQ1QCqlQnr2DhmGtsJeTW4ivk4VX9CHY8Cot8b86Viw/bY3HlB+0UblDYvPcuwaipxdzF7ckZMEXz6DGEwN8GUo1clz+niCP8/k461OyVdXCaYq8hpReh3Ct99ge2A8xoDPm/G3QIAGHcHa3tgnFP4SF4RbBF66cOfg0aTDDJ23SAKik7VJBSqHsXh6DKpqeGjVrZQEji5UUxK3cG5EyDnXKA91N7oHBDw8pUe0CLO3MatPKZkTQl3qcddr6ZO0esKLqXVDsX6RtK6/GI9LYbG/r498R0svMURV/D5w735ROw6k03EzFaKA8dl1NLM4i4zj06T1j7gfz4hVbttXo1ONQo1kc4BeP3MXYkNtiUV480VcWsrg24DGT6VltovHLgVc62a8LI60kMb4duRIkIOvYPEyK4x4itYo1Fm/da/AEvWvDXDiRtnmWl6zcSv5DKyzuSsLZCPNqmEGJ2WSCOh6yfEUKd6Tzoy8wsDO+P+RG6iL5KWaZBP/xmZJW0VmJwNKedvowlPKu44ZlagQ07if+CPpKkKphWqb+5rBaulpQiMK/AhVvq8X6x6g1dp8pYaYE0hUg4uFS2B21W4cB3SCUdilP+/MymkHjCexx9K/zCsVMw7CxIE/mZGr0AN0eQa5eIRryZXbObBtBKRExXqzT7kjaCc6a9AC0ByQV35yShui/QgrdEsM15fdJjFy1Skj8rydXbVAcKWe8OML3bGrHuPCWDH3vVNHBT3+xeCwj7Vw/G3Onu8VeoSpfulmsIdzBO3xFOqCEgUmyew+EVR7nihm4Q+PxFBsbShn4zg1WhhDkfsNtwYEH5sFX4Q1U6woB+kBqR2eDL9+6Zb9QzXYLpTGdi/8w5oPsGkiJjEB3hWbzIUQ1XMDo6GzRE3swx9dldJIUOX/kM6dJSzTlM7aZ1qJJB65g0zbssc4l5dqhPxkKDxY9Hd/eSDfsoicpppuYbxQ0SqqQ3GbRa4T+TchckSZPaIc3mEPBKgQLZkCBJ9Z4vD+FrGIB38LP+ItKQZ1PzGTEG0HlsxypoQc+93dYLEPzcHvaoWTRBDWk0uMpOFPY6Y3TGvdvGVwbp920EH6K4CiiObmnFNCaSDi9L8ym+sSynR48mJ7BSMz//7nAAieeHOiLcuAevjGMJUIXhmYHQdDW/9VXnof+I1uRc4/YwFwM9auKdCIx9Hv9I7OUmmP19i9U1f9BproY0kV41qkFfVj0mxMzDgQL8Bfr/VcsBTCGp0i7aojQ/XQohlMLMSO8szb4FLmQAaOXVUonHTzYHu8wL4JBSHqEx6qBqFfI5AkDXImftwrT+QpbiSiTxYE1DLRdIur/MmOSQbRo7qDdfueetx6tH4DJ/Yl2+4mvOY7kqYNDNpEk18PVoRfJU7bWb+0SiQfQriVkNecESa876m+g8gTiuYWBYyby2vRVt2HsmrSC1A33xOLOvda0y9x7+jiil07z36reylgWKkmhbHJ/vv/rYLwocV50x1JKqIkxWchSoHUdqpE6ueVP6IXX+m3rxMdNC2tyy1g2QQa2RRDBDBYrpzZjfSGQFJf8Nyifiuir6fSZ+xJelBQZr2/KXj/C8LyR6JMXvN8zo1Y8B4+M6bp0FzZCCM5LbXenO5ZunX/YqvYRjqu4B8xQJ2m5t2i2cguJp4qvF8o8VlHVSGnkKGtoGkpxcg/J/rY0pc/4aKgIisdT96BCn3sIEMUugqTtasW1K0KaIbTNlfG0TpXVjjLX7RyKSFD2Xac2FA2N38VgpEe71gpZzxfhezr8XkqaskRIrR49MGZFrk7oCICPuXRkdeiv6VpYYU9zbMJXqsulStyOVTjgjEGiYQAvVx/M+fLxX7xKygzzA0ZTh/3VlzSYCvKTkIxfIidFY+JEHcJzwG9DsGVT5fAjIujj8CXZB5YNYrGpp0FKWoqRi60ebRIZEAMI8GWn0soX0A6d+la0zF+z/H5IJFDq5kFIzgAUpjdMcPa7YjevZlZq5AFM8g1r1g/HfQEfZ48rM+VUW38HyvY82Qw2df2Df6hvGr2jJZqRKWLGdF+zvsAkiP/nlvAi/4Pz+sjOg8aecGMenW3u8DNYyyjlrBMLn95RU0JzLh+XJGbwTBJFCZ9nA17LkaF+s7/iXdGxZpbayzJBtZ3fJK7fiaIUyvrd12z262tBXw6JOwnsC9MEqWzYKztEk9JXKIl0jeDwotz0axzeFeBdMtjBRsilTeiXbARqf2rMJwiPU9GRfIyZNFjKxbf1fxAbN2ggMX0koo/xBG8ghLvNV2TUdHMo9vNUiIx1xOJx3GcJBIa+/C0nhzOLwqo1Ylyqc5VaNghD3lfmPCp40yTaw02IBL0NL0bNsYS0pk0rnn+Rr229L1GrNuon1UKaI1I42L2HLTgvZI8D8S4tOQhk6FYM+Wtpzek9/faHkmmWZ2LwfgWZarARBjLMfmrXtlX3gjh+KI6Y/7oMudyauALHNQ7ZCRugaphL92aYYz/4v1Nin/wFB4ybZNGxHE/a2jy0HaWUmB3LMLrNiP7+vFg1orX/1GWQqARUyHFodzJAGuobMqdUk3qLQ6bbcs9gdp68iBK8cx1HOvtVvteCj+vv9JaCsmQihJ/dNYIo4Rf3UvdnDQONkaO4O2rRTfvQ9LCjWClBXBel0BxcJ9HOU12oV4SBHkKDYvTb05KGmPRfm/lKLD0XjUuaVF0EA5b53PKlE65TMZuGjcgY0L+rPKYXYRilDKORBBAdEKfwjSUNH6S/Ds3TCXanXmJTTsdHq1Q1DZEQgI6yUA5QT/jX9Mam3SPDCHXEkjFeY6tQ6t3XJ1UrZSQ1u2SVFQtHAkNAWtWfdpSM/puAGTwwAsG8Du9RZjOpQPwX8fEV19NqFSTrgITigTBKewh9W6Iii3qWMqiHrP31Q6jsGWc1MkJh71/AhwFM6WSg8G9GxwEwHHgE7J65YWFXgrRKyB+bD2+aNKsY+xUPDyHU62Cz21uVTVSlWB1d28QzbsrVEp9m4SZL4PfkVM/lpVDuyOMk6+snn/0Ss+rDUgz2QLEx4VheNoUHMKmEwpg+I2yOOO1Gn4B7+SC4SB+GTy2qf9x81vdlzClyjuhVbiuOnc+ErgAEAkoeI2biehX+VezyairjJw5KvfN0EV9eWyGqOPH81tgEbV7UJJIARhc6XWopU+8Iei1ekSzGJulWsAvasgv9FwaBAdA61GooAT3RryqL9s/wqSp0/OA7OAcGvuPvyaAi7z4K5rMkJRspu6hoMGp0ia/7+ceB0QoUd7yVftT83QHufg7oubHSnsraOSLyu3/Pcndj7OKIiZILBoH1v2hTFPWjQ0qbWgbcHx9W9h9um3+4Qc2Ru6a4KcGtgZNb+XwSnF+ssCN99C8nZ7w+3JLEhf48KEKBDQLGGe/kylwQucrZ0Bi0CNGAK+kx7M10yuBMhqrYTqMWs2YMsYDdHCezhLRVcAHlciuGPJDFvAF7hze5H23MkGocdjXbk0S4OxkAuL7sAZw+C+cDg/rtkQvxp22aeMpGaJZ+qmx75yPiH8NepR0kqCm2xEh644n3ysSVg92P0fDs5KOOVr+VWIPKL7BFFHfwG7PiE9KxA3cOIygWMRBwp3SEHbXXNr85tOUdmXRN2PIajU4GweOC7/f3YCm2ZBGe0kDzKMTAaFjoOx1YnpSwZZZGCbncd96h83SUJ1IwO8th/UI3yVsPKiiCQEYjvYvrc0X2NcUthXObWfMbtu13aVeGlxN5HqhYPlpctSVsSX0yYrPUWOidV7z/8P+sBjcQv5MW/Cr57F8qNRETfDOlB1/RL3sagXqtnYIlHJtBx0LJaqyZHMYcfYBTZiWRy6qa1aK6xqT2HyAiXkLg3kB/T4m7EYffeX/9iQ3+aFVW7uHuUCZq+rheV0Hl3IGDsjjMpxBUlFk9Dn/+AcZYPZBIWfvt+oYy51pCEZHnprpBx6kmKOpKMkoBPtJKX1EadLCypxSGi1pbomlz1NxXfWqDTNVcO1ILTli8CwA9oMHzmN8yHHlkc2nHMR1qDPSXTTvFLlZ1RBQ0WznOjczZc67nrqYKzfjJbO0NxVgh6HmAuOLGb+OI/2er+oAR2tNVUPrzWFp/gMxJTCpIQXfU/IbvOIyy59O26woLhcsmsmH451/OpX56U6wypvrIoin7Zl+AsMyQGBnCSQQvg4uGS59Xi/nnK8w3BYHfpGxBHAe1UsPhAJ5MDNGKtJJf3+qh3cZFqPe/7nWlNvh4F3jm1WwOhctI8gybbXQ4Lcr/q3JqH1zlfeWAW8hNNYMCjxyIA966fg7CZfk5xKIcXrIbp4vg7wDHDlYw1hDhkwC7zldwLqsgonr/a4rBGdslAISE1Y43mgZ6s/K+Y4ResrzoihGjXHs0HqQNWPvRKYAJmevbPm8V7AlaifW/liTQn82t/Ezlyqp8kKUP02EUrcGKGc8kOSa52jI0cieeFBzhbR3oItCp7cpqt63odJnxITX2RGAqMMMyhP7fkm9m8oyjqV2Frh/Ty0I6O0IKtvTTfM8JyCAmKm1iEprWsUn1Ofvd2t9A4ofEz8r/fqGpmXtjzsCvibmvDtJT77wtHPtZMdeE2R20dZ5L0U2kAn/dzZD2jSsePO4Moz7hvVV7nMnXUtMullY3Nf0dyrH5eoWVW3xxi93wKuTRVay8fzylyWhYu0IoxpVNDNvO1JiuT5HK5wcbWRS6rr1q7hzRqYjUowoL74/FIHSniwJeSI2D5yCIxj9q1SqVGAojRFXwdQGkENrUPd45BclQkLO76xFnfY5Gcavyv5/hE8SIPWMDUYDOx7OgI1UgpPetAgeQyBJvIjhJgCPzB7y7l8QHW2myvWmXpe2l/0NHAm2fUPbvyz2Ojz58abXFw9pOiq+pMhx95ocklg1aTXtcgL2Zc8GAM+9HyzD4cFx2ubeN9gVwVpC3gXPWwbmnY8O7ZrQpiS2UyGziBkqR/4OgvtA10WSZkD3t0RJrBIEzMADLCMWr+WtMMpswX9GwJZdEA/+oeruX4b23RMGG+m5JJTMpbpTKkIu2a6MvpIxeY5/Jm3Poumv7Z7CFPbo5B5dnqyPfoVrCvjUYN01PBMbSQszv7v3dVtlpEHFxrsbxEYxnQe6Lb1qxsvw24QUSJJw5cHxyHwfBsfdRCYqXNNJN7HU5EWmsFiMTqxSZ1pg/QQszLRCE7oz/LLs/8Z3oeWcfZAGpPqCVNkeayaMgn7KhHybk9/sXbU8ISrcN+zIWKn/DVbDkKWedPUYAQ=="
  }
]
#7 · Assistant
2026-06-30 21:47:13
Model · claude-sonnet-5
Thinking
I'm working on Addendum D to address the "Self-gauging, attempted safely" thread by introducing the Paired Witness Protocol—a concrete procedure with a visual artifact. The idea is to create a twinned gauge or ritual diagram as an SVG object that captures how calibration drift propagates recursively through the surveyor's own measurements.

The core problem I'm tackling: if the surveyor's own drift shifts their readings, then two surveyors gauging each other simultaneously create a feedback loop. I'm designing a solution using sequential gauging rather than simultaneous—alternating turns with a built-in lag—and introducing a "Witness Token" (a stamped disc each surveyor carries) that gets gauged first thing each morning to establish their personal offset for the day, which then gets subtracted from all subsequent readings. I'm also sketching out a "Twinning Plate" as a paired ledger sheet for recording these sequential self and other readings, plus a "Kinship Assay Card" procedure for reverse-engineering unknown gauges by testing them against known springs in a triangulation-like pattern to narrow down their casting source.

For Addendum D, I'm focusing the Witness Coin protocol as the centerpiece—it directly solves the self-gauging safety problem—and keeping the kinship assay as a lighter secondary touch. The Witness Coin itself has nice thematic resonance: cast from the surveyor's own birth-water or birth-related material (playing on the "caul" pun), it becomes a personal null reference parallel to how the Sill Spring serves as the null reference for objects. Each morning the surveyor gauges their Coin first using the full three-anchor method to establish their daily offset, then subtracts that constant from all other readings taken that day.

I'm also introducing a Paired Protocol for two surveyors working together—they trade Coins and gauge each other's instrument sequentially rather than simultaneously, keeping the readings separate and comparing afterward. I'll add an SVG diagram showing the Coin with casting marks and recast schedule, plus a simple ASCII log table format for daily tracking.

The data reveals something interesting: some surveyors' Coins read with negative current offset, meaning they're more bound than usual. This ties back to the Negative Rim Bound concept—maybe veteran surveyors who've taken oaths of service to the Survey itself naturally register as bound, which would explain why negative offsets cluster among experienced practitioners. That opens threads for Addendum E about whether paired surveyors' Coins drift toward each other over time through repeated work together, creating a kind of work-kinship effect.

I'm also flagging open questions: what happens if a Coin is lost or destroyed, and what does a newborn's caul-cast Coin actually read at birth—is it truly zero, or does it show something unexpected compared to adult recasts?

Writing the full section now with the same flat bureaucratic field-log voice, adding SVG diagrams for the Witness Coin and paired protocol, keeping the brass and wax aesthetic consistent with the survey lettering style.
# FIELD INSTRUMENT NO. 1 — ADDENDUM D ### The Witness Coin, and a Protocol for Safe Self-Gauging (submitted through unofficial channels; the courier has started signing these "filed, not approved") The open question that wouldn't wait: if the gauge reads relative to a living Witness, and the Witness has Drift of their own, every reading in the permanent record is contaminated by whoever was standing at the fourth anchor that day. You cannot subtract an unknown. But you can measure it, separately, before it touches anything else — provided you stop trying to gauge a person with the same instrument that needs that person to function. The fix did not come from the Instruments Office. It came from noting that the gauge already has a precedent for a personal null reference. It's in the name. A caul is not a generic membrane — it is the membrane, the one each person sheds once, at the start, before they have had any chance to drift at all. --- ## I. The Witness Coin A small flat disc, cast — where the relic survives — from a surveyor's own caul, sealed in the same anchor-wax as the gauge's bob, and never carried in the same hand as a working instrument. Where no caul-relic survives (most cases; they are not commonly kept), the Coin is cast instead at the surveyor's formal naming, witnessed and dated, and stands as a declared zero rather than a born one — a distinction the Survey insists on recording, though it has yet to find a case where the two read differently. The Coin is gauged like any other target — full three anchors, Door/Name/Flame, by another person if one is present, or alone if not, since the Coin is not the surveyor and the collapse that ruined the Named-Twice attempts does not apply to it. The resulting number is not the surveyor's Drift. It is the Personal Offset — how much of today's reading was contributed by the hand on the housing, before that hand touches anything else.
WITNESS COIN cast: caul-relic, where extant else: naming-day, witnessed Handling rule Never carried in the gauge-hand. Never gauged by the person it represents. Gauged fresh each working morning, before any other target is touched. A Coin read only once and reused all season is not an Offset. It is a guess wearing the shape of a number. Reading on Coin = Personal Offset for the day. Subtract from, or flag against, all targets gauged after. The Offset is not the surveyor's Drift. It is only the part of the surveyor that leaked into the reading.
``` --- ## II. The Daily Offset Log A working week, transcribed from a field copy (surveyor's initials withheld at her request):
DATE     COIN READING   BAND/SIGN      NOTE
3.iv     1.1 Cauls      Tide           ordinary morning, slept badly
4.iv     0.4 Cauls      Still          —
5.iv     −1.6 Cauls     Bound          witnessed a colleague's oath-renewal at breakfast,
                                       unclear if relevant, flagged anyway
6.iv     0.9 Cauls      Tide           —
7.iv     −2.3 Cauls     Bound          (see Open Threads — this is the third Bound
                                       reading this surveyor has logged this season)
``` A Tide-band Offset is treated as background noise and ignored. A Wane or worse, logged on the Coin, means the day's readings on everything else should be marked provisional regardless of how clean the targets' own anchors were — the Witness brought too much weather into the room. This has already overturned one prior finding: a Husk-band reading on a storage-room door, taken on a day its surveyor's Coin read 5.8 Wane, has been struck from the permanent record and re-walked. The door, re-gauged on a clean day, came back Tide. The drift had been the surveyor's, on loan to the door. --- ## III. The Paired Protocol For two surveyors who must gauge each other's Drift directly — not their Coins, but the people — the rule is now: never simultaneously.
A B Step 1 A gauges B. B is Target, A is Witness. Log sealed, unread by B. Step 2 — separately, after a pause B gauges A. A is Target, B is Witness. Log sealed, unread by A. Step 3 Both logs opened together. Each Coin-Offset already on file is subtracted before comparing.
``` The pause between steps is mandatory, not ceremonial: a same-day double-reading, run back to back, was tried once at Bench Sill and produced two numbers seven Cauls apart with no agreement on which surveyor had drifted into the other's reading. The transcript is in the appendix and resolves nothing. The Survey's current position is that simultaneous mutual gauging may simply not be a coherent operation — like asking a single bob to settle toward two centers at once. It doesn't refuse, the way Named-Twice refuses. It just produces noise shaped enough like data to be dangerous. --- ## IV. A Note on the Bound Readings Three surveyors so far have logged Coins that read consistently negative — Kept or Bound, never Still, never Tide. All three have given long service. The Instruments Office wants this filed as equipment fault. The surveyors involved have, independently, declined to have their Coins recast. --- ## Open Threads (for Addendum E) - Does the Offset compound? Two surveyors who work together for a season, each logging mild Tide-band Coins daily, gauged each other at season's end and found Bound-band readings neither had shown alone. Is shared work a vow the gauge can detect before the people involved would call it one? - The newborn Coin. No caul-relic Coin has ever been gauged in its first season, while the surveyor it represents is still an infant. If true zero exists anywhere, this is the likeliest candidate. Nobody has been willing to make the attempt with full anchors on a child too young to meet the Name anchor for themselves. - What does it mean to refuse a recast? Three veteran Coins, all Bound, all declined for re-casting by their owners. Is declining itself a Bound act — a vow renewed by the refusal to reset it? - The kinship-assay, still undone. Unstamped gauges remain in the field, and the Witness Coin protocol does nothing to identify a gauge's birth-water. Someone still has to run an unstamped instrument pairwise against three known springs and triangulate its casting-source from the bias pattern alone, the way a stranger's accent gives away a country.
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "I'm working on Addendum D to address the \"Self-gauging, attempted safely\" thread by introducing the Paired Witness Protocol—a concrete procedure with a visual artifact. The idea is to create a twinned gauge or ritual diagram as an SVG object that captures how calibration drift propagates recursively through the surveyor's own measurements.\n\nThe core problem I'm tackling: if the surveyor's own drift shifts their readings, then two surveyors gauging each other simultaneously create a feedback loop. I'm designing a solution using sequential gauging rather than simultaneous—alternating turns with a built-in lag—and introducing a \"Witness Token\" (a stamped disc each surveyor carries) that gets gauged first thing each morning to establish their personal offset for the day, which then gets subtracted from all subsequent readings. I'm also sketching out a \"Twinning Plate\" as a paired ledger sheet for recording these sequential self and other readings, plus a \"Kinship Assay Card\" procedure for reverse-engineering unknown gauges by testing them against known springs in a triangulation-like pattern to narrow down their casting source.\n\nFor Addendum D, I'm focusing the Witness Coin protocol as the centerpiece—it directly solves the self-gauging safety problem—and keeping the kinship assay as a lighter secondary touch. The Witness Coin itself has nice thematic resonance: cast from the surveyor's own birth-water or birth-related material (playing on the \"caul\" pun), it becomes a personal null reference parallel to how the Sill Spring serves as the null reference for objects. Each morning the surveyor gauges their Coin first using the full three-anchor method to establish their daily offset, then subtracts that constant from all other readings taken that day.\n\nI'm also introducing a Paired Protocol for two surveyors working together—they trade Coins and gauge each other's instrument sequentially rather than simultaneously, keeping the readings separate and comparing afterward. I'll add an SVG diagram showing the Coin with casting marks and recast schedule, plus a simple ASCII log table format for daily tracking.\n\nThe data reveals something interesting: some surveyors' Coins read with negative current offset, meaning they're more bound than usual. This ties back to the Negative Rim Bound concept—maybe veteran surveyors who've taken oaths of service to the Survey itself naturally register as bound, which would explain why negative offsets cluster among experienced practitioners. That opens threads for Addendum E about whether paired surveyors' Coins drift toward each other over time through repeated work together, creating a kind of work-kinship effect.\n\nI'm also flagging open questions: what happens if a Coin is lost or destroyed, and what does a newborn's caul-cast Coin actually read at birth—is it truly zero, or does it show something unexpected compared to adult recasts?\n\nWriting the full section now with the same flat bureaucratic field-log voice, adding SVG diagrams for the Witness Coin and paired protocol, keeping the brass and wax aesthetic consistent with the survey lettering style.",
    "signature": "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"
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM D\n### *The Witness Coin, and a Protocol for Safe Self-Gauging*\n*(submitted through unofficial channels; the courier has started signing these \"filed, not approved\")*\n\nThe open question that wouldn't wait: if the gauge reads relative to a living Witness, and the Witness has Drift of their own, every reading in the permanent record is contaminated by whoever was standing at the fourth anchor that day. You cannot subtract an unknown. But you can *measure* it, separately, before it touches anything else — provided you stop trying to gauge a person with the same instrument that needs that person to function.\n\nThe fix did not come from the Instruments Office. It came from noting that the gauge already has a precedent for a personal null reference. It's in the name. A **caul** is not a generic membrane — it is *the* membrane, the one each person sheds once, at the start, before they have had any chance to drift at all.\n\n---\n\n## I. The Witness Coin\n\nA small flat disc, cast — where the relic survives — from a surveyor's own caul, sealed in the same anchor-wax as the gauge's bob, and never carried in the same hand as a working instrument. Where no caul-relic survives (most cases; they are not commonly kept), the Coin is cast instead at the surveyor's formal naming, witnessed and dated, and stands as a *declared* zero rather than a *born* one — a distinction the Survey insists on recording, though it has yet to find a case where the two read differently.\n\nThe Coin is gauged like any other target — full three anchors, Door/Name/Flame, by *another* person if one is present, or alone if not, since the Coin is not the surveyor and the collapse that ruined the Named-Twice attempts does not apply to it. The resulting number is not the surveyor's Drift. It is the **Personal Offset** — how much of *today's* reading was contributed by the hand on the housing, before that hand touches anything else.\n\n```svg\n<svg viewBox=\"0 0 420 280\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect width=\"420\" height=\"280\" fill=\"#f1e9d8\"/>\n\n  <!-- the coin itself -->\n  <circle cx=\"110\" cy=\"110\" r=\"70\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\" stroke-width=\"2.5\"/>\n  <circle cx=\"110\" cy=\"110\" r=\"58\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1\"/>\n  <!-- caul-membrane texture, irregular concentric tear-lines -->\n  <path d=\"M60 95 Q90 70 130 80 Q160 90 155 120 Q150 150 115 150 Q75 150 65 125 Q58 110 60 95 Z\"\n        fill=\"none\" stroke=\"#7a1f1f\" stroke-width=\"1\" opacity=\"0.6\"/>\n  <text x=\"110\" y=\"115\" font-size=\"11\" text-anchor=\"middle\" fill=\"#3a2e1f\">WITNESS</text>\n  <text x=\"110\" y=\"130\" font-size=\"11\" text-anchor=\"middle\" fill=\"#3a2e1f\">COIN</text>\n  <text x=\"110\" y=\"195\" font-size=\"9\" text-anchor=\"middle\" fill=\"#5c3b1e\">cast: caul-relic, where extant</text>\n  <text x=\"110\" y=\"208\" font-size=\"9\" text-anchor=\"middle\" fill=\"#5c3b1e\">else: naming-day, witnessed</text>\n\n  <!-- the recast schedule, like the gauge's own seasonal rule -->\n  <g transform=\"translate(230,40)\" font-size=\"10\" fill=\"#3a2e1f\">\n    <text x=\"0\" y=\"0\" font-weight=\"bold\">Handling rule</text>\n    <text x=\"0\" y=\"20\">Never carried in the gauge-hand.</text>\n    <text x=\"0\" y=\"36\">Never gauged by the person it represents.</text>\n    <text x=\"0\" y=\"52\">Gauged fresh each working morning,</text>\n    <text x=\"0\" y=\"66\">before any other target is touched.</text>\n    <text x=\"0\" y=\"86\" font-style=\"italic\" fill=\"#7a1f1f\">A Coin read only once and reused</text>\n    <text x=\"0\" y=\"100\" font-style=\"italic\" fill=\"#7a1f1f\">all season is not an Offset. It is a guess</text>\n    <text x=\"0\" y=\"114\" font-style=\"italic\" fill=\"#7a1f1f\">wearing the shape of a number.</text>\n  </g>\n\n  <line x1=\"20\" y1=\"235\" x2=\"400\" y2=\"235\" stroke=\"#3a2e1f\" stroke-width=\"0.5\" stroke-dasharray=\"1 3\"/>\n  <text x=\"20\" y=\"255\" font-size=\"10\" fill=\"#3a2e1f\">\n    Reading on Coin = Personal Offset for the day. Subtract from, or flag against, all targets gauged after.\n  </text>\n  <text x=\"20\" y=\"270\" font-size=\"9\" font-style=\"italic\" fill=\"#5c3b1e\">\n    The Offset is not the surveyor's Drift. It is only the part of the surveyor that leaked into the reading.\n  </text>\n</svg>\n```\n\n---\n\n## II. The Daily Offset Log\n\nA working week, transcribed from a field copy (surveyor's initials withheld at her request):\n\n```\nDATE     COIN READING   BAND/SIGN      NOTE\n3.iv     1.1 Cauls      Tide           ordinary morning, slept badly\n4.iv     0.4 Cauls      Still          —\n5.iv     −1.6 Cauls     Bound          witnessed a colleague's oath-renewal at breakfast,\n                                       unclear if relevant, flagged anyway\n6.iv     0.9 Cauls      Tide           —\n7.iv     −2.3 Cauls     Bound          (see Open Threads — this is the third Bound\n                                       reading this surveyor has logged this season)\n```\n\nA Tide-band Offset is treated as background noise and ignored. A Wane or worse, logged on the Coin, means the day's readings on *everything else* should be marked provisional regardless of how clean the targets' own anchors were — the Witness brought too much weather into the room. This has already overturned one prior finding: a Husk-band reading on a storage-room door, taken on a day its surveyor's Coin read 5.8 Wane, has been struck from the permanent record and re-walked. The door, re-gauged on a clean day, came back Tide. The drift had been the surveyor's, on loan to the door.\n\n---\n\n## III. The Paired Protocol\n\nFor two surveyors who must gauge each other's Drift directly — not their Coins, but the people — the rule is now: **never simultaneously.**\n\n```svg\n<svg viewBox=\"0 0 460 200\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect width=\"460\" height=\"200\" fill=\"#f1e9d8\"/>\n\n  <g transform=\"translate(40,40)\">\n    <circle cx=\"20\" cy=\"20\" r=\"18\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <text x=\"20\" y=\"24\" text-anchor=\"middle\" font-size=\"10\">A</text>\n  </g>\n  <g transform=\"translate(360,40)\">\n    <circle cx=\"20\" cy=\"20\" r=\"18\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <text x=\"20\" y=\"24\" text-anchor=\"middle\" font-size=\"10\">B</text>\n  </g>\n\n  <text x=\"230\" y=\"55\" text-anchor=\"middle\" font-weight=\"bold\">Step 1</text>\n  <line x1=\"80\" y1=\"60\" x2=\"380\" y2=\"60\" stroke=\"#7a1f1f\" stroke-width=\"1.5\" marker-end=\"url(#arrow)\"/>\n  <text x=\"230\" y=\"80\" text-anchor=\"middle\" font-size=\"10\">A gauges B. B is Target, A is Witness. Log sealed, unread by B.</text>\n\n  <text x=\"230\" y=\"115\" text-anchor=\"middle\" font-weight=\"bold\">Step 2 — separately, after a pause</text>\n  <line x1=\"380\" y1=\"120\" x2=\"80\" y2=\"120\" stroke=\"#7a1f1f\" stroke-width=\"1.5\" marker-end=\"url(#arrow)\"/>\n  <text x=\"230\" y=\"140\" text-anchor=\"middle\" font-size=\"10\">B gauges A. A is Target, B is Witness. Log sealed, unread by A.</text>\n\n  <text x=\"230\" y=\"170\" text-anchor=\"middle\" font-weight=\"bold\">Step 3</text>\n  <text x=\"230\" y=\"186\" text-anchor=\"middle\" font-size=\"10\">Both logs opened together. Each Coin-Offset already on file is subtracted before comparing.</text>\n\n  <defs>\n    <marker id=\"arrow\" markerWidth=\"8\" markerHeight=\"8\" refX=\"6\" refY=\"4\" orient=\"auto\">\n      <path d=\"M0 0 L8 4 L0 8 Z\" fill=\"#7a1f1f\"/>\n    </marker>\n  </defs>\n</svg>\n```\n\nThe pause between steps is mandatory, not ceremonial: a same-day double-reading, run back to back, was tried once at Bench Sill and produced two numbers seven Cauls apart with no agreement on which surveyor had drifted into the other's reading. The transcript is in the appendix and resolves nothing. The Survey's current position is that simultaneous mutual gauging may simply not be a coherent operation — like asking a single bob to settle toward two centers at once. It doesn't refuse, the way Named-Twice refuses. It just produces noise shaped enough like data to be dangerous.\n\n---\n\n## IV. A Note on the Bound Readings\n\nThree surveyors so far have logged Coins that read consistently negative — Kept or Bound, never Still, never Tide. All three have given long service. The Instruments Office wants this filed as equipment fault. The surveyors involved have, independently, declined to have their Coins recast.\n\n---\n\n## Open Threads (for Addendum E)\n\n- **Does the Offset compound?** Two surveyors who work together for a season, each logging mild Tide-band Coins daily, gauged each other at season's end and found Bound-band readings neither had shown alone. Is shared work a vow the gauge can detect before the people involved would call it one?\n- **The newborn Coin.** No caul-relic Coin has ever been gauged in its first season, while the surveyor it represents is still an infant. If true zero exists anywhere, this is the likeliest candidate. Nobody has been willing to make the attempt with full anchors on a child too young to meet the Name anchor for themselves.\n- **What does it mean to refuse a recast?** Three veteran Coins, all Bound, all declined for re-casting by their owners. Is declining itself a Bound act — a vow renewed by the refusal to reset it?\n- **The kinship-assay, still undone.** Unstamped gauges remain in the field, and the Witness Coin protocol does nothing to identify a gauge's birth-water. Someone still has to run an unstamped instrument pairwise against three *known* springs and triangulate its casting-source from the bias pattern alone, the way a stranger's accent gives away a country."
  }
]
#8 · Assistant
2026-06-30 21:50:16
Model · claude-sonnet-5
Thinking
I'm working on Addendum E now, pulling together the established framework—cauls, bands, anchors with their four types (Door, Name, Flame, Witness), the Witness Coin, Personal Offset, and the Sill Springs tied to Site 19's glass kinship-bias—and weaving in the Negative Rim concepts (Kept, Bound, Named-Twice), Cross-Gauge mechanics, and how kinship-assay can be undone. I'll create visual artifacts and structured tables to make these interconnections clear and specific.

I'm settling on two main threads for this section: first, the kinship assay itself—a triangulation method using an unstamped gauge tested against three known springs (Coast, Bench, Glass) to map its birth-water composition and reveal mixed provenance in gauges; second, the compounding Offset problem when two surveyors work together and develop a Bound reading neither could achieve alone, introducing concepts like Shared Caul or Wedlock Drift as a new collaborative phenomenon. I'll anchor this with a ternary plot diagram showing how an unknown gauge's offset pattern triangulates to its origin, then build supporting tables and prose, leaving open threads about newborn Coins and proxy witnessing for the next addendum. Creating SVG diagrams now...
# FIELD INSTRUMENT NO. 1 — ADDENDUM E ### The Kinship Assay, and the Yoke Question (courier's note, verbatim, retained on the envelope: "filed, not approved, but I notice nobody's told me to stop") Two threads converged this season that the Instruments Office did not ask to be related, and are anyway. The first is mechanical: how do you read a gauge's birth-water when nobody wrote it down. The second is human: what happens when two people's Offsets stop adding and start agreeing. --- ## I. The Kinship Assay The method follows from Addendum C directly: if an unwitnessed gauge favors its own birth-water in a Cross-Gauge, then the pattern of favoritism is a signature, not noise. Run an unstamped gauge against three springs of known provenance, witnessed properly each time so the bias isn't masked by the Fourth Anchor doing its job too well, and plot the three resulting offsets against each other. A gauge cast purely from one spring will sit hard in a corner. A gauge cast from blended wax — which the foundry ledger admits happened more than once, in lean years, when one spring's water ran short mid-casting — will sit somewhere in the field between.
Site 19 — Glass Sill Site 4 — Coast Sill Site 11 — Bench Sill Gauge No. 233 62% / 24% / 14%(?) Gauge No. 118 98% Site 19 — clean cast No. 233's three readings sum to 86%, not 100%. The missing 14% does not belong to any known spring. Kinship Assay, ternary plot — composition by Cross-Gauge bias against three known sources
``` Gauge No. 233 was chosen for the demonstration because it was the first to fail the arithmetic. Its three pairwise biases, properly witnessed, do not sum to a full account. Fourteen parts in a hundred point at nothing in the ledger. Two more unstamped gauges, assayed since, show the same shortfall, in different proportions, never zero. The Instruments Office's preferred explanation is measurement error compounding across three separate trials. This does not explain why the shortfall is consistent in direction — all three gauges, on the same sub-axis of the plot, lean the same way toward the same absent corner, as if there is a fourth vertex the triangle simply isn't drawn wide enough to show. Nobody has proposed a name for it. The foundry ledger, consulted again, lists exactly three sources for the relevant casting years. Either the ledger is wrong, or there was a fourth spring in use that nobody thought worth recording — which, given how casually Site 19 itself went unremarked-upon for ninety years, is not a comfortable thing to rule out. --- ## II. The Yoke Question Meanwhile, away from the foundry archive entirely, two surveyors closed their season's logs and did the comparison Addendum D recommends — Coin against Coin, each day's Personal Offset subtracted before any joint reading was trusted — and found something the subtraction does not explain.
SURVEYOR M — Daily Coin (selected)          SURVEYOR R — Daily Coin (selected)
wk 1   0.8 Tide                              wk 1   1.0 Tide
wk 6   0.6 Tide                              wk 6   0.5 Tide
wk 12  0.7 Tide                              wk 12  0.6 Tide
wk 18  0.9 Tide                              wk 18  0.4 Tide

— individually, neither surveyor logged a single Wane-or-worse Coin all season —

SEASON-END PAIRED GAUGE, M reading R, then R reading M (full protocol, paused, sealed separately):
    M → R :  −2.2 Cauls   (Bound)
    R → M :  −2.4 Cauls   (Bound)

Predicted by simple subtraction of season-average Offsets: ~1.4 Cauls combined, Tide-band, unremarkable.
Actual: both Bound. Both surveyors. Independently sealed. Opened together.
``` Subtraction assumes the two Offsets are unrelated noise contributed separately to a shared reading. The result says otherwise: whatever these two surveyors had been doing all season — and the log notes only "shared a circuit, six sites, no incident" — had produced something in both of them, detectable only when they pointed the gauge at each other, invisible to either Coin alone. The working name in the field copy is the Yoke: a Bound-grade commitment that exists between two people rather than in either one, undetectable by any instrument that only ever measures one end of a pair at a time. The Witness Coin, built to isolate a single person's contribution, cannot see it by design — there is no single Coin to read. A modification was attempted before the season closed: a single bead of wax, drawn half from M's Coin-casting and half from R's, melted together rather than laid side by side, and gauged as its own target with a third surveyor as Witness.
Coin M Coin R Yoke bead Yoke bead, gauged alone: −2.3 Cauls — Bound consistent with the paired cross-reading, taken directly, no subtraction required. The Yoke bead reads the same Bound value the paired protocol found by indirection. It does not explain the Yoke. It only confirms it has a single, measurable location.
``` The bead works, in the narrow sense that it produces a stable, repeatable number instead of the noise a simultaneous mutual gauging gives. It does not tell anyone what the Yoke is — vow, habit, accumulated mutual witnessing, or something the gauge has no better word for than the one it already has for everything else it measures: a thing that has stopped being two separate things and started being recognizably, measurably, itself. Neither M nor R has asked to have the bead destroyed. Both have asked, separately, not to be told the exact number again. --- ## Open Threads (for Addendum F) - The fourth vertex. Three unstamped gauges now show the same directional shortfall in the Kinship Assay. If a fourth, unrecorded spring contributed wax to the foundry during the lean years, it has never been located, and every "clean" single-source gauge cast in that window may be quietly part-Yoked to a place with no name in any ledger. - Yoke beads, plural. Nobody has tried a three-way bead. Whether the Yoke is strictly a pairwise property or scales to crews, families, or entire survey teams is untested, and the Instruments Office has expressed, in writing, that it would "prefer this not be tested on Office time." - Declining the bead. M and R's refusal to hear the number again sits next to the three veteran surveyors' refusal to recast their Bound Coins (Addendum D). A pattern is forming: every confirmed negative reading so far has been met not with correction but with protection. Is refusal-to-recast and refusal-to-be-told the same gesture, performed by the instrument's subjects rather than its operators? - The newborn Coin, still untried. Still open. Still nobody willing to attempt it. - A small thing from the assay margins: Gauge No. 233's missing 14%, plotted against No. 118's clean 98% Site 19, fall on roughly the same bearing from the Glass Sill vertex — as though the unnamed fourth source and Site 19 are, themselves, somehow related. Two open questions may be one question wearing two coats.
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "I'm working on Addendum E now, pulling together the established framework—cauls, bands, anchors with their four types (Door, Name, Flame, Witness), the Witness Coin, Personal Offset, and the Sill Springs tied to Site 19's glass kinship-bias—and weaving in the Negative Rim concepts (Kept, Bound, Named-Twice), Cross-Gauge mechanics, and how kinship-assay can be undone. I'll create visual artifacts and structured tables to make these interconnections clear and specific.\n\nI'm settling on two main threads for this section: first, the kinship assay itself—a triangulation method using an unstamped gauge tested against three known springs (Coast, Bench, Glass) to map its birth-water composition and reveal mixed provenance in gauges; second, the compounding Offset problem when two surveyors work together and develop a Bound reading neither could achieve alone, introducing concepts like Shared Caul or Wedlock Drift as a new collaborative phenomenon. I'll anchor this with a ternary plot diagram showing how an unknown gauge's offset pattern triangulates to its origin, then build supporting tables and prose, leaving open threads about newborn Coins and proxy witnessing for the next addendum. Creating SVG diagrams now...",
    "signature": "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"
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM E\n### *The Kinship Assay, and the Yoke Question*\n*(courier's note, verbatim, retained on the envelope: \"filed, not approved, but I notice nobody's told me to stop\")*\n\nTwo threads converged this season that the Instruments Office did not ask to be related, and are anyway. The first is mechanical: how do you read a gauge's birth-water when nobody wrote it down. The second is human: what happens when two people's Offsets stop adding and start *agreeing*.\n\n---\n\n## I. The Kinship Assay\n\nThe method follows from Addendum C directly: if an unwitnessed gauge favors its own birth-water in a Cross-Gauge, then the *pattern* of favoritism is a signature, not noise. Run an unstamped gauge against three springs of **known** provenance, witnessed properly each time so the bias isn't masked by the Fourth Anchor doing its job too well, and plot the three resulting offsets against each other. A gauge cast purely from one spring will sit hard in a corner. A gauge cast from blended wax — which the foundry ledger admits happened more than once, in lean years, when one spring's water ran short mid-casting — will sit somewhere in the field between.\n\n```svg\n<svg viewBox=\"0 0 460 420\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect width=\"460\" height=\"420\" fill=\"#f1e9d8\"/>\n\n  <!-- ternary triangle -->\n  <polygon points=\"230,40 80,330 380,330\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2.5\"/>\n\n  <!-- internal gridlines, 25/50/75 -->\n  <g stroke=\"#9a8a6e\" stroke-width=\"0.75\" stroke-dasharray=\"2 3\">\n    <line x1=\"155\" y1=\"185\" x2=\"305\" y2=\"185\"/>\n    <line x1=\"117\" y1=\"257\" x2=\"343\" y2=\"257\"/>\n    <line x1=\"192\" y1=\"112\" x2=\"268\" y2=\"112\"/>\n    <line x1=\"155\" y1=\"185\" x2=\"155\" y2=\"185\"/>\n  </g>\n\n  <!-- vertex labels -->\n  <text x=\"230\" y=\"28\" text-anchor=\"middle\" font-weight=\"bold\">Site 19 — Glass Sill</text>\n  <text x=\"80\" y=\"350\" text-anchor=\"middle\" font-weight=\"bold\">Site 4 — Coast Sill</text>\n  <text x=\"380\" y=\"350\" text-anchor=\"middle\" font-weight=\"bold\">Site 11 — Bench Sill</text>\n\n  <!-- example plot: Gauge No. 233 -->\n  <circle cx=\"245\" cy=\"195\" r=\"6\" fill=\"#7a1f1f\" stroke=\"#3a2e1f\"/>\n  <text x=\"255\" y=\"190\" font-size=\"10\">Gauge No. 233</text>\n  <text x=\"255\" y=\"204\" font-size=\"9\" fill=\"#5c3b1e\">62% / 24% / 14%(?)</text>\n\n  <!-- a clean single-source gauge for contrast -->\n  <circle cx=\"230\" cy=\"58\" r=\"6\" fill=\"#3a2e1f\"/>\n  <text x=\"240\" y=\"62\" font-size=\"10\">Gauge No. 118</text>\n  <text x=\"240\" y=\"76\" font-size=\"9\" fill=\"#5c3b1e\">98% Site 19 — clean cast</text>\n\n  <!-- the residual problem -->\n  <text x=\"80\" y=\"380\" font-size=\"10\" font-style=\"italic\" fill=\"#7a1f1f\">\n    No. 233's three readings sum to 86%, not 100%.\n  </text>\n  <text x=\"80\" y=\"396\" font-size=\"10\" font-style=\"italic\" fill=\"#7a1f1f\">\n    The missing 14% does not belong to any known spring.\n  </text>\n\n  <text x=\"230\" y=\"412\" text-anchor=\"middle\" font-size=\"9\" fill=\"#5c3b1e\">\n    Kinship Assay, ternary plot — composition by Cross-Gauge bias against three known sources\n  </text>\n</svg>\n```\n\nGauge No. 233 was chosen for the demonstration because it was the first to fail the arithmetic. Its three pairwise biases, properly witnessed, do not sum to a full account. Fourteen parts in a hundred point at *nothing in the ledger.* Two more unstamped gauges, assayed since, show the same shortfall, in different proportions, never zero.\n\nThe Instruments Office's preferred explanation is measurement error compounding across three separate trials. This does not explain why the shortfall is consistent in *direction* — all three gauges, on the same sub-axis of the plot, lean the same way toward the same absent corner, as if there is a fourth vertex the triangle simply isn't drawn wide enough to show.\n\nNobody has proposed a name for it. The foundry ledger, consulted again, lists exactly three sources for the relevant casting years. Either the ledger is wrong, or there was a fourth spring in use that nobody thought worth recording — which, given how casually Site 19 itself went unremarked-upon for ninety years, is not a comfortable thing to rule out.\n\n---\n\n## II. The Yoke Question\n\nMeanwhile, away from the foundry archive entirely, two surveyors closed their season's logs and did the comparison Addendum D recommends — Coin against Coin, each day's Personal Offset subtracted before any joint reading was trusted — and found something the subtraction does not explain.\n\n```\nSURVEYOR M — Daily Coin (selected)          SURVEYOR R — Daily Coin (selected)\nwk 1   0.8 Tide                              wk 1   1.0 Tide\nwk 6   0.6 Tide                              wk 6   0.5 Tide\nwk 12  0.7 Tide                              wk 12  0.6 Tide\nwk 18  0.9 Tide                              wk 18  0.4 Tide\n\n— individually, neither surveyor logged a single Wane-or-worse Coin all season —\n\nSEASON-END PAIRED GAUGE, M reading R, then R reading M (full protocol, paused, sealed separately):\n    M → R :  −2.2 Cauls   (Bound)\n    R → M :  −2.4 Cauls   (Bound)\n\nPredicted by simple subtraction of season-average Offsets: ~1.4 Cauls combined, Tide-band, unremarkable.\nActual: both Bound. Both surveyors. Independently sealed. Opened together.\n```\n\nSubtraction assumes the two Offsets are unrelated noise contributed separately to a shared reading. The result says otherwise: whatever these two surveyors had been doing all season — and the log notes only \"shared a circuit, six sites, no incident\" — had produced something in *both of them*, detectable only when they pointed the gauge at each other, invisible to either Coin alone.\n\nThe working name in the field copy is the **Yoke**: a Bound-grade commitment that exists between two people rather than in either one, undetectable by any instrument that only ever measures one end of a pair at a time. The Witness Coin, built to isolate a single person's contribution, cannot see it by design — there is no single Coin to read.\n\nA modification was attempted before the season closed: a single bead of wax, drawn half from M's Coin-casting and half from R's, melted together rather than laid side by side, and gauged as its own target with a third surveyor as Witness.\n\n```svg\n<svg viewBox=\"0 0 440 220\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect width=\"440\" height=\"220\" fill=\"#f1e9d8\"/>\n\n  <circle cx=\"140\" cy=\"100\" r=\"55\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"140\" y=\"40\" text-anchor=\"middle\" font-size=\"10\">Coin M</text>\n  <circle cx=\"240\" cy=\"100\" r=\"55\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"240\" y=\"40\" text-anchor=\"middle\" font-size=\"10\">Coin R</text>\n\n  <!-- overlap zone, melted bead -->\n  <path d=\"M190 55 A55 55 0 0 1 190 145 A55 55 0 0 1 190 55 Z\" fill=\"#e3d7bd\" opacity=\"0.9\" stroke=\"#7a1f1f\" stroke-width=\"1.5\"/>\n  <text x=\"190\" y=\"103\" text-anchor=\"middle\" font-size=\"9\" fill=\"#7a1f1f\">Yoke</text>\n  <text x=\"190\" y=\"116\" text-anchor=\"middle\" font-size=\"9\" fill=\"#7a1f1f\">bead</text>\n\n  <text x=\"340\" y=\"60\" font-size=\"11\" font-weight=\"bold\">Yoke bead, gauged alone:</text>\n  <text x=\"340\" y=\"80\" font-size=\"13\" fill=\"#7a1f1f\">−2.3 Cauls — Bound</text>\n  <text x=\"340\" y=\"100\" font-size=\"9\" fill=\"#5c3b1e\">consistent with the paired</text>\n  <text x=\"340\" y=\"113\" font-size=\"9\" fill=\"#5c3b1e\">cross-reading, taken directly,</text>\n  <text x=\"340\" y=\"126\" font-size=\"9\" fill=\"#5c3b1e\">no subtraction required.</text>\n\n  <text x=\"20\" y=\"190\" font-size=\"10\" font-style=\"italic\">\n    The Yoke bead reads the same Bound value the paired protocol found by indirection.\n  </text>\n  <text x=\"20\" y=\"206\" font-size=\"10\" font-style=\"italic\">\n    It does not explain the Yoke. It only confirms it has a single, measurable location.\n  </text>\n</svg>\n```\n\nThe bead works, in the narrow sense that it produces a stable, repeatable number instead of the noise a *simultaneous* mutual gauging gives. It does not tell anyone what the Yoke *is* — vow, habit, accumulated mutual witnessing, or something the gauge has no better word for than the one it already has for everything else it measures: a thing that has stopped being two separate things and started being recognizably, measurably, itself.\n\nNeither M nor R has asked to have the bead destroyed. Both have asked, separately, not to be told the exact number again.\n\n---\n\n## Open Threads (for Addendum F)\n\n- **The fourth vertex.** Three unstamped gauges now show the same directional shortfall in the Kinship Assay. If a fourth, unrecorded spring contributed wax to the foundry during the lean years, it has never been located, and every \"clean\" single-source gauge cast in that window may be quietly part-Yoked to a place with no name in any ledger.\n- **Yoke beads, plural.** Nobody has tried a three-way bead. Whether the Yoke is strictly a pairwise property or scales to crews, families, or entire survey teams is untested, and the Instruments Office has expressed, in writing, that it would \"prefer this not be tested on Office time.\"\n- **Declining the bead.** M and R's refusal to hear the number again sits next to the three veteran surveyors' refusal to recast their Bound Coins (Addendum D). A pattern is forming: every confirmed negative reading so far has been met not with correction but with *protection*. Is refusal-to-recast and refusal-to-be-told the same gesture, performed by the instrument's subjects rather than its operators?\n- **The newborn Coin, still untried.** Still open. Still nobody willing to attempt it.\n- **A small thing from the assay margins**: Gauge No. 233's missing 14%, plotted against No. 118's clean 98% Site 19, fall on roughly the same bearing from the Glass Sill vertex — as though the unnamed fourth source and Site 19 are, themselves, somehow related. Two open questions may be one question wearing two coats."
  }
]
#9 · Assistant
2026-06-30 21:53:18
Model · claude-sonnet-5
Thinking
Now I'm planning Addendum F, picking up the threads that have been left hanging—the newborn Coin that's been marked "still open" multiple times, the fourth vertex that needs a concrete identity and location tied to Site 19, and the three-way bead test that could either resolve or deepen the mystery. I'll structure it around finally committing to these plot points rather than deferring them again. The newborn Coin attempt finally happens—a family lets them try it on an infant, but since the child hasn't been named yet, they can't complete the Name anchor, so they proceed with an incomplete reading anyway. The needle doesn't settle into any known Band; instead it oscillates, fluttering around center without deciding—a new phenomenon I'm calling "UNSET," where the gauge itself hasn't determined what stillness means yet. It's a poetic resolution: a newborn's identity genuinely hasn't solidified, so the instrument reflects that pre-Band oscillation.

For the fourth vertex, I'm tying it to temporal rather than spatial identity—the missing 14% of castings correlates with a specific historical event at Site 19, a "Long Tide" or lean year when the spring itself was reportedly "thin" or dormant. This suggests the fourth vertex isn't a missing location but Site 19 at a different time, implying that Sill Springs themselves have temporal identity and can drift across years, which loops back to the original question about whether a spring can shift its nature.

I'm also planning a lighter third section on saturation: a three-way bead test with a family shows the Yoke reading caps at the same depth regardless of how many people are bound, suggesting the instrument has a measurable limit where it can't distinguish strength beyond a certain point.

Writing SVG diagrams...
# FIELD INSTRUMENT NO. 1 — ADDENDUM F ### The Newborn Coin, and the Fourth Spring's Shadow (the courier no longer brings an envelope; she copies these by hand and leaves before anyone can ask her to stop) Two surveyors made the attempt nobody else would, on the same week, by coincidence, in different districts. And the foundry's missing fourteen percent turned out not to be a place at all. Both findings reach back to the very first open question this log ever recorded — whether a Sill Spring can drift — and the answer, when it finally arrived, was worse than a yes. --- ## I. The Newborn Coin Surveyor R — half of the pair who produced last season's Yoke bead — had a niece born in the spring. The family consented, against the Office's written discomfort, to a single attempt: a caul-relic Coin, cast within the hour of birth, gauged the same evening with full Door and Flame anchors met and Witness held by R herself, since the child obviously could not hold her own. The Name anchor could not be met. The child had not yet been named — the family follows the old custom of waiting a season — and R would not borrow the name under consideration without permission. The gauge was run anyway, anchors incomplete, in the spirit of the provisional tildes this whole log has been carrying since Vask's wedding ring. The needle did not settle. Not slowly, not provisionally — it did not arrive anywhere at all. Across eleven minutes of held reading it traced a wide, repeating fan, never twice in quite the same arc, never once still long enough to read off a number.
0 / true Still eleven minutes, untranscribable as a single number field note: "the needle was not undecided. It was busy." — Surveyor R, naming-eve trial
``` This is not Vask's tilde, and it is not the noise the simultaneous mutual-gauging trial produced at Bench Sill. That noise looked like a number lying. This looked like no number had been decided yet to lie about. The field copy proposes a new entry for the glyph table, placed not at either end of the scale but athwart it, since it does not belong to the positive Bands or the negative ones:
QUICK no Band yet proposed, unranked — the gauge has no word for becoming, only for having become.
``` The family declined a second attempt. The naming, when it happens this autumn, will not be followed by a re-gauging. R's note: "I don't think the number was ever going to be the point. I think we wanted to know if she'd hold still, and she answered honestly." --- ## II. The Fourth Spring's Shadow The foundry ledger has a gap nobody flagged until someone went looking specifically for it: a six-year span, the so-called lean years, when Site 19's recorded output dips sharply and casting records become sloppy, abbreviated, sometimes just a tally mark and a date. Three of the unstamped gauges with the persistent 14% shortfall were all cast inside this window. None outside it shows the shortfall at all. The assumption since Addendum E was a fourth, unlisted spring. A second look at the lean-years entries — water volume, not water source — suggested a different question: what if Site 19 didn't run short because of a drought, but because something about the spring itself changed, mid-decade, and the foundry kept calling it Site 19 out of habit while casting from water that was not, by the Kinship Assay's own logic, quite the same substance anymore? A retired foundry hand, tracked down through the lean-years ledger's barely-legible signature, still had two sealed sample-jars of raw casting wax from before and after the gap. Both were run through the Kinship Assay against three known references, including modern Site 19.
Site 19 (modern) Site 4 Site 11 jar, pre-gap — 95% Site 19 jar, post-gap — only 81% toward same unnamed bearing as Gauge No. 233's shortfall drift, same spring, six years apart Two jars. One spring, by every ledger and every label. Two different chemistries of stillness.
``` Both jars are labeled Site 19, in the same hand, eleven years apart. The Assay says they are not quite the same water. The "fourth source" three unstamped gauges seem to share a trace of is not a hidden spring at all — it is Site 19 before whatever happened to it, drifted just far enough from its current self that a gauge can tell the two apart, though no ledger ever thought to. This closes the very first open thread this log carried, from before there was a log: can a Sill Spring itself drift, and has any null reference been re-tested past a century. The answer, on the evidence of two sealed jars, is yes — not past a century, but within one working man's career, and nobody noticed because nobody had a second jar to check it against until he happened to have kept one. It also means every "clean, single-source" gauge cast from Site 19 across ninety years may be reading against a target that has not stayed in one place the whole time it was being used to define place. The Instruments Office's official position, delivered verbally and never put in writing, was that this finding "is not being filed at this time." It is being filed here instead. --- ## III. A Smaller Note: The Bead Does Not Add Before the season closed, three colleagues — not a couple, a survey crew of long standing — agreed to a three-way Yoke bead, melted from equal parts of each Coin, gauged by a fourth, uninvolved Witness.
Pairwise beads (built first, for comparison):
   A+B  —  −2.1 Cauls   Bound
   A+C  —  −1.9 Cauls   Bound
   B+C  —  −2.4 Cauls   Bound

Three-way bead, A+B+C melted together:
   −2.3 Cauls   Bound
``` A three-person commitment did not read deeper than a two-person one. It read into the same narrow band the pairs already occupied, as if Bound has a floor the gauge cannot resolve past from this side — the mirror image, perhaps, of Named-Twice fusing solid on the far end of the positive scale. The instrument may simply lack the resolution to say whether three people are yoked more than two, only that they are yoked, full stop, past a certain depth, the same way it cannot say how much more than zero Vacant is. --- ## Open Threads (for Addendum G) - QUICK is not yet a Band. It has no anchors, no provisional sign, no place on either rim. Does it belong on the instrument at all, or is it evidence that the instrument was only ever built to read what has already finished arriving — and a newborn, by definition, hasn't? - If Site 19 drifted once, it can drift again. Nobody has proposed a re-test schedule for any spring, ever, in ninety years of Cataloguing. The lean-years jars survived by accident, not policy. What is the Survey's plan for telling the difference between a spring and a moving target it has been mistaking for one? - The Bound floor. If Bound saturates past a certain depth regardless of headcount, is there a better instrument that could resolve it — or is "more bound than the gauge can say" simply what the deepest Band feels like from outside, the same way Vacant looks identical whether something is barely gone or has been gone for years? - The retired foundry hand kept his two jars for eleven years without being asked to, for no stated reason. He has more jars. He has not yet said how many, or from where, or whether any of them predate the ledger entirely.
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "Now I'm planning Addendum F, picking up the threads that have been left hanging—the newborn Coin that's been marked \"still open\" multiple times, the fourth vertex that needs a concrete identity and location tied to Site 19, and the three-way bead test that could either resolve or deepen the mystery. I'll structure it around finally committing to these plot points rather than deferring them again. The newborn Coin attempt finally happens—a family lets them try it on an infant, but since the child hasn't been named yet, they can't complete the Name anchor, so they proceed with an incomplete reading anyway. The needle doesn't settle into any known Band; instead it oscillates, fluttering around center without deciding—a new phenomenon I'm calling \"UNSET,\" where the gauge itself hasn't determined what stillness means yet. It's a poetic resolution: a newborn's identity genuinely hasn't solidified, so the instrument reflects that pre-Band oscillation.\n\nFor the fourth vertex, I'm tying it to temporal rather than spatial identity—the missing 14% of castings correlates with a specific historical event at Site 19, a \"Long Tide\" or lean year when the spring itself was reportedly \"thin\" or dormant. This suggests the fourth vertex isn't a missing location but Site 19 at a different time, implying that Sill Springs themselves have temporal identity and can drift across years, which loops back to the original question about whether a spring can shift its nature.\n\nI'm also planning a lighter third section on saturation: a three-way bead test with a family shows the Yoke reading caps at the same depth regardless of how many people are bound, suggesting the instrument has a measurable limit where it can't distinguish strength beyond a certain point.\n\nWriting SVG diagrams...",
    "signature": "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"
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM F\n### *The Newborn Coin, and the Fourth Spring's Shadow*\n*(the courier no longer brings an envelope; she copies these by hand and leaves before anyone can ask her to stop)*\n\nTwo surveyors made the attempt nobody else would, on the same week, by coincidence, in different districts. And the foundry's missing fourteen percent turned out not to be a place at all. Both findings reach back to the very first open question this log ever recorded — whether a Sill Spring can drift — and the answer, when it finally arrived, was worse than a yes.\n\n---\n\n## I. The Newborn Coin\n\nSurveyor R — half of the pair who produced last season's Yoke bead — had a niece born in the spring. The family consented, against the Office's written discomfort, to a single attempt: a caul-relic Coin, cast within the hour of birth, gauged the same evening with full Door and Flame anchors met and Witness held by R herself, since the child obviously could not hold her own.\n\nThe Name anchor could not be met. The child had not yet been named — the family follows the old custom of waiting a season — and R would not borrow the name under consideration without permission. The gauge was run anyway, anchors incomplete, in the spirit of the provisional tildes this whole log has been carrying since Vask's wedding ring.\n\nThe needle did not settle. Not slowly, not provisionally — it did not arrive anywhere at all. Across eleven minutes of held reading it traced a wide, repeating fan, never twice in quite the same arc, never once still long enough to read off a number.\n\n```svg\n<svg viewBox=\"0 0 380 320\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect width=\"380\" height=\"320\" fill=\"#f1e9d8\"/>\n  <circle cx=\"190\" cy=\"150\" r=\"120\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2.5\"/>\n  <line x1=\"190\" y1=\"30\" x2=\"190\" y2=\"45\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"190\" y=\"22\" text-anchor=\"middle\" font-size=\"12\" fill=\"#3a2e1f\">0 / true Still</text>\n\n  <!-- fan of needle traces, faint, multiple positions, none settled -->\n  <g stroke=\"#7a1f1f\" stroke-width=\"1\" opacity=\"0.5\">\n    <line x1=\"190\" y1=\"150\" x2=\"140\" y2=\"55\"/>\n    <line x1=\"190\" y1=\"150\" x2=\"245\" y2=\"50\"/>\n    <line x1=\"190\" y1=\"150\" x2=\"100\" y2=\"95\"/>\n    <line x1=\"190\" y1=\"150\" x2=\"280\" y2=\"100\"/>\n    <line x1=\"190\" y1=\"150\" x2=\"165\" y2=\"45\"/>\n    <line x1=\"190\" y1=\"150\" x2=\"215\" y2=\"42\"/>\n    <line x1=\"190\" y1=\"150\" x2=\"120\" y2=\"135\"/>\n    <line x1=\"190\" y1=\"150\" x2=\"265\" y2=\"150\"/>\n  </g>\n  <circle cx=\"190\" cy=\"150\" r=\"4\" fill=\"#3a2e1f\"/>\n\n  <text x=\"190\" y=\"265\" text-anchor=\"middle\" font-size=\"11\" fill=\"#3a2e1f\">eleven minutes, untranscribable as a single number</text>\n  <text x=\"190\" y=\"283\" text-anchor=\"middle\" font-size=\"10\" fill=\"#7a1f1f\" font-style=\"italic\">field note: \"the needle was not undecided. It was busy.\"</text>\n  <text x=\"190\" y=\"300\" text-anchor=\"middle\" font-size=\"9\" fill=\"#5c3b1e\">— Surveyor R, naming-eve trial</text>\n</svg>\n```\n\nThis is not Vask's tilde, and it is not the noise the simultaneous mutual-gauging trial produced at Bench Sill. That noise looked like a number lying. This looked like *no number had been decided yet to lie about.*\n\nThe field copy proposes a new entry for the glyph table, placed not at either end of the scale but athwart it, since it does not belong to the positive Bands or the negative ones:\n\n```svg\n<svg viewBox=\"0 0 200 90\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect width=\"200\" height=\"90\" fill=\"#f1e9d8\"/>\n  <g transform=\"translate(20,15)\">\n    <circle cx=\"20\" cy=\"20\" r=\"14\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n    <line x1=\"20\" y1=\"20\" x2=\"8\" y2=\"8\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n    <line x1=\"20\" y1=\"20\" x2=\"32\" y2=\"6\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n    <line x1=\"20\" y1=\"20\" x2=\"6\" y2=\"24\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n    <line x1=\"20\" y1=\"20\" x2=\"34\" y2=\"28\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n    <line x1=\"20\" y1=\"20\" x2=\"22\" y2=\"2\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n    <text x=\"20\" y=\"50\" text-anchor=\"middle\">QUICK</text>\n    <text x=\"20\" y=\"64\" text-anchor=\"middle\" font-size=\"9\">no Band yet</text>\n  </g>\n  <text x=\"80\" y=\"30\" font-size=\"10\" width=\"100\">proposed, unranked —</text>\n  <text x=\"80\" y=\"44\" font-size=\"10\">the gauge has no</text>\n  <text x=\"80\" y=\"58\" font-size=\"10\">word for becoming,</text>\n  <text x=\"80\" y=\"72\" font-size=\"10\">only for having become.</text>\n</svg>\n```\n\nThe family declined a second attempt. The naming, when it happens this autumn, will not be followed by a re-gauging. R's note: *\"I don't think the number was ever going to be the point. I think we wanted to know if she'd hold still, and she answered honestly.\"*\n\n---\n\n## II. The Fourth Spring's Shadow\n\nThe foundry ledger has a gap nobody flagged until someone went looking specifically for it: a six-year span, the so-called **lean years**, when Site 19's recorded output dips sharply and casting records become sloppy, abbreviated, sometimes just a tally mark and a date. Three of the unstamped gauges with the persistent 14% shortfall were all cast inside this window. None outside it shows the shortfall at all.\n\nThe assumption since Addendum E was a fourth, unlisted *spring*. A second look at the lean-years entries — water volume, not water source — suggested a different question: what if Site 19 didn't run short because of a drought, but because something about the spring itself changed, mid-decade, and the foundry kept calling it Site 19 out of habit while casting from water that was not, by the Kinship Assay's own logic, quite the same substance anymore?\n\nA retired foundry hand, tracked down through the lean-years ledger's barely-legible signature, still had two sealed sample-jars of raw casting wax from before and after the gap. Both were run through the Kinship Assay against three known references, including modern Site 19.\n\n```svg\n<svg viewBox=\"0 0 460 300\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect width=\"460\" height=\"300\" fill=\"#f1e9d8\"/>\n\n  <polygon points=\"230,30 80,250 380,250\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"230\" y=\"18\" text-anchor=\"middle\" font-size=\"10\" font-weight=\"bold\">Site 19 (modern)</text>\n  <text x=\"80\" y=\"266\" text-anchor=\"middle\" font-size=\"10\" font-weight=\"bold\">Site 4</text>\n  <text x=\"380\" y=\"266\" text-anchor=\"middle\" font-size=\"10\" font-weight=\"bold\">Site 11</text>\n\n  <!-- pre-gap sample, sits hard near Site 19 vertex -->\n  <circle cx=\"225\" cy=\"55\" r=\"6\" fill=\"#3a2e1f\"/>\n  <text x=\"245\" y=\"52\" font-size=\"9\">jar, pre-gap — 95% Site 19</text>\n\n  <!-- post-gap sample, displaced toward unmarked interior point -->\n  <circle cx=\"245\" cy=\"150\" r=\"6\" fill=\"#7a1f1f\"/>\n  <text x=\"255\" y=\"148\" font-size=\"9\">jar, post-gap — only 81%</text>\n  <text x=\"255\" y=\"161\" font-size=\"9\" fill=\"#7a1f1f\">toward same unnamed bearing</text>\n  <text x=\"255\" y=\"174\" font-size=\"9\" fill=\"#7a1f1f\">as Gauge No. 233's shortfall</text>\n\n  <!-- line connecting them, drift vector -->\n  <line x1=\"225\" y1=\"55\" x2=\"245\" y2=\"150\" stroke=\"#7a1f1f\" stroke-width=\"1.5\" stroke-dasharray=\"3 3\"/>\n  <text x=\"190\" y=\"105\" font-size=\"9\" font-style=\"italic\" fill=\"#7a1f1f\" transform=\"rotate(78 190 105)\">drift, same spring, six years apart</text>\n\n  <text x=\"80\" y=\"290\" font-size=\"10\" font-style=\"italic\">\n    Two jars. One spring, by every ledger and every label. Two different chemistries of stillness.\n  </text>\n</svg>\n```\n\nBoth jars are labeled Site 19, in the same hand, eleven years apart. The Assay says they are not quite the same water. The \"fourth source\" three unstamped gauges seem to share a trace of is not a hidden spring at all — it is **Site 19 before whatever happened to it**, drifted just far enough from its current self that a gauge can tell the two apart, though no ledger ever thought to.\n\nThis closes the very first open thread this log carried, from before there was a log: *can a Sill Spring itself drift, and has any null reference been re-tested past a century.* The answer, on the evidence of two sealed jars, is yes — not past a century, but within one working man's career, and nobody noticed because nobody had a second jar to check it against until he happened to have kept one.\n\nIt also means every \"clean, single-source\" gauge cast from Site 19 across ninety years may be reading against a target that has not stayed in one place the whole time it was being used to define *place.* The Instruments Office's official position, delivered verbally and never put in writing, was that this finding \"is not being filed at this time.\" It is being filed here instead.\n\n---\n\n## III. A Smaller Note: The Bead Does Not Add\n\nBefore the season closed, three colleagues — not a couple, a survey crew of long standing — agreed to a three-way Yoke bead, melted from equal parts of each Coin, gauged by a fourth, uninvolved Witness.\n\n```\nPairwise beads (built first, for comparison):\n   A+B  —  −2.1 Cauls   Bound\n   A+C  —  −1.9 Cauls   Bound\n   B+C  —  −2.4 Cauls   Bound\n\nThree-way bead, A+B+C melted together:\n   −2.3 Cauls   Bound\n```\n\nA three-person commitment did not read deeper than a two-person one. It read into the same narrow band the pairs already occupied, as if Bound has a floor the gauge cannot resolve past from this side — the mirror image, perhaps, of Named-Twice fusing solid on the far end of the positive scale. The instrument may simply lack the resolution to say whether three people are yoked *more* than two, only that they are yoked, full stop, past a certain depth, the same way it cannot say how much more than zero Vacant is.\n\n---\n\n## Open Threads (for Addendum G)\n\n- **QUICK is not yet a Band.** It has no anchors, no provisional sign, no place on either rim. Does it belong on the instrument at all, or is it evidence that the instrument was only ever built to read what has already finished arriving — and a newborn, by definition, hasn't?\n- **If Site 19 drifted once, it can drift again.** Nobody has proposed a re-test schedule for any spring, ever, in ninety years of Cataloguing. The lean-years jars survived by accident, not policy. What is the Survey's plan for telling the difference between a spring and a moving target it has been mistaking for one?\n- **The Bound floor.** If Bound saturates past a certain depth regardless of headcount, is there a *better* instrument that could resolve it — or is \"more bound than the gauge can say\" simply what the deepest Band feels like from outside, the same way Vacant looks identical whether something is barely gone or has been gone for years?\n- **The retired foundry hand** kept his two jars for eleven years without being asked to, for no stated reason. He has more jars. He has not yet said how many, or from where, or whether any of them predate the ledger entirely."
  }
]
#10 · Assistant
2026-06-30 21:56:21
Model · claude-sonnet-5
Thinking
I'm picking up the threads from the previous addendum—QUICK as a non-band, the drifting spring problem, the Bound floor—and now focusing on the retired foundry hand's jar collection as a secret archive. I want to introduce new SVG artifacts that feel discovered, develop the mechanical and conceptual ideas further, and leave open threads for the next section.

The foundry hand's jars become a chronological "Drift Almanac"—a shelf diagram spanning decades showing continuous position drift at Site 19 over time, plotted on the ternary diagram to reveal cyclical or trending movement. For QUICK, I'm proposing it reads things still in their becoming phase, which opens up a new instrument design called "the Loom" or "the Quickening Gauge"—something that tracks rate of change directly rather than settling position, a fundamentally different mechanism. And for the Bound floor, I'm connecting it elegantly to Site 19's pre-gap position: the floor's numerical value matches the drift vector magnitude, tying together the asymmetric readings and creating a resonant motif across the whole system.

Now I'm structuring Addendum G with four main sections: the Foundry Hand's shelf revealing decades of drift, the Quickening Gauge as a new instrument design, the numeric echo between Bound floor and Site 19 drift, and a teaser about an even older unlabeled jar that opens threads for Addendum H. I'll need to create SVG diagrams for the shelf, the new gauge schematic, and an overlay showing the numeric correspondence, then write it all in the same archival tone with courier notes and open threads at the end. The Wax Ceiling concept ties together the extremes—Named-Twice as positive saturation and Bound floor as negative saturation, both expressions of the same material limit. Now I'm expanding the foundry hand's archive to reveal decades of private tracking predating the official lean years, suggesting he was running his own long experiment before anyone else noticed the drift. I'll include a transcribed excerpt from his ledger in his own hand with cryptic entries, then open threads about whether the Wax Ceiling is a material property or a conceptual one, and whether Site19's drift follows a cycle with implications for the Survey's calibration itself.

I'm also introducing the Quickening Gauge as a test of rate-based measurement on something other than a newborn, and the mysterious unlabeled jar from the foundry hand's pre-ledger period that doesn't match any known spring—possibly the missing fourth vertex or something stranger. Now I'll write the full addendum with consistent voice and styling, keeping the SVGs valid and readable with the established palette.
# FIELD INSTRUMENT NO. 1 — ADDENDUM G ### The Foundry Hand's Shelf, and the Quickening Gauge (the courier left this one inside a jar; the Office returned the jar, empty, with no comment) The retired foundry hand — let the log finally use his name, since he has stopped objecting: Coen Ashery, forty-one years at the Site 19 works, eleven of them after retirement spent doing something nobody asked him to do. He had more jars. He had a great many more jars. And somewhere in sorting them, the Bound floor and the Newborn fan turned out to be reaching for the same wall from opposite sides. --- ## I. The Almanac Ashery's shelf, when finally inventoried, held thirty-eight sealed sample-jars of raw Site 19 casting wax, dated continuously back well before the lean years anyone else had noticed — the earliest jar predates the foundry ledger's own first legible entry by nine years. He kept one jar roughly every fourteen months, unprompted, "because the water never poured the same way twice and somebody should write that down even if it isn't anybody's job." Run through the Kinship Assay against the same three references, the jars do not show a single drift event. They show a cycle.
Ashery's shelf, 38 jars, ~14 months apart — red rims mark "lean" readings % match to MODERN Site 19, by jar date 100% 75% lean trough, yr ~14 second trough, yr ~33 period ≈ 19 years, not yet closed twice
``` The "lean years" were not an anomaly. They were the trough of a period Ashery's own shelf catches twice, roughly nineteen years apart, and may be catching a third time now, partway down the second descent. The Third Cataloguing calibrated true Still against Site 19 at a point on this curve nobody marked as a point on a curve, because nobody suspected there was a curve to be caught at a phase of. Every reading the Survey has ever taken is phase-locked to wherever Site 19 happened to be sitting on a cycle whose existence postdates the instrument that depends on it. Ashery's note, copied from the inside lid of the newest jar, lid-side, where he apparently expected nobody would look until after he'd stopped being askable about it: > "I am not smart enough to know if it breathes or just wobbles on its bedrock the way anything heavy does over enough years. I only know it does not hold still, and a spring that does not hold still has no business being the thing that tells everyone else whether they do." --- ## II. The Quickening Gauge Addendum F left QUICK off the scale entirely — a fan, not a point, eleven minutes of motion that never arrived. The standard instrument cannot read it because the standard instrument was built to report where, and the newborn's needle was reporting how fast, which is a different question the old housing has no dial for. A new instrument, built rather than adapted: not a needle over a fixed dial, but a stylus over a slow-turning drum, recording the spread of the fan continuously instead of forcing it to a single settled point.
drum turns at fixed rate — band-width = QUICK reading stylus, free to swing — never forced to settle Quicken Index = band-width, in degrees-of-arc per minute of recording a narrowing band over successive turns is read as "settling" — the gauge's first instrument capable of detecting that an object is in the act of arriving. R's niece, re-run on the Quickening Gauge at three months: band narrowing, not yet closed.
``` R's niece was the first and so far only trial subject, gauged again at three months with the family's permission, no Name anchor still required since the instrument doesn't ask for a settled point. The band has narrowed by roughly a third since the naming-eve session. It has not closed. The working guess, unconfirmed, is that it closes somewhere near the time a child would ordinarily be expected to answer to a name on her own, without it being offered to her — which would mean QUICK isn't a Band so much as a countdown to the moment Name becomes available as an anchor at all. --- ## III. The Wax Ceiling The coincidence Addendum F left lying in plain sight: the deepest Bound readings across every Yoke bead trial cluster around −2.3 to −2.4 Cauls and refuse to go lower no matter how many people are melted into the bead. Ashery's pre-gap jar, run for the first time against itself — that is, against a contemporary Site 19 sample, the closest thing to gauging a spring's drift directly against its own younger water — returns a number from the same neighborhood.
−3 0 the Ceiling all Yoke-bead floors 2-way, 3-way alike pre-gap jar vs modern Site 19 Both lines of evidence saturate at the same wax-physical limit, not the same conceptual one.
``` The working theory, stated plainly for the first time: Bound has no floor of its own. What the gauge reports as "deepest possible commitment" may simply be the deepest strain the anchor-wax can hold before its molecular memory of position stops discriminating — the same physical ceiling, approached from the other direction, that fuses Named-Twice's bob solid past +3. Vacant and the floor of Bound may be the same wall, read from opposite faces, and the instrument has never been built that could tell the difference between "this relation is functionally infinite" and "this relation is merely past where my wax can still keep score." If true, every Bound reading this log has ever logged as a number was never a number. It was a flag: deeper than I can say. The three veteran surveyors' declined recasts, M and R's bead, the three-way crew — none of them may differ from each other at all. The gauge has been reporting a ceiling and calling it a measurement. --- ## IV. The Jar With No Date Among Ashery's thirty-eight, one jar carries no label, no date, and — unlike every other jar on the shelf — no trace of Site 19 wax chemistry whatsoever. Run against all three known references and against Ashery's own oldest dated Site 19 sample, it matches none of them, including itself: two assays of the same jar's contents, taken eight days apart by the same operator with full anchors met, returned different ternary positions. Ashery's only comment, given to the surveyor who found it and transcribed faithfully: "That one I didn't pour. I found it already on the shelf when I started, in '04, older lid-wax than anything we were casting that decade. I kept meaning to ask the old foundry master where it came from. He died before I worked up the nerve, and I have been too much of a coward to open it again since the second reading came back wrong twice in a row." --- ## Open Threads (for Addendum H) - The cycle's third trough. If the ~19-year period holds, the Survey is partway down a second descent right now, with no record of anyone having calibrated anything against the previous trough — meaning the current "true Still" may be sliding, slowly, under every reading taken this decade, undetected because nobody had a second jar to check it against until Ashery. - The Ceiling, tested properly. Does the wax-physical saturation limit move if the anchor-wax is cast from a different spring entirely — Coast, Bench, even the unnamed fourth bearing from the Assay margins? If the Ceiling is material rather than relational, a non-Site-19 Yoke bead should saturate at a different number. Nobody has built one. - The Quickening Gauge, on adults. Tried only on an infant so far. Does a freshly sworn oath, a newly dug well, a marriage an hour old, show the same narrowing band — meaning QUICK isn't reserved for the literally newborn, but for anything that hasn't finished arriving yet, on any scale? - The undated jar. Two assays, same contents, two different answers. Either the assay failed, the jar's contents are actively changing between readings, or — least comfortable, most consistent with everything else this season has found — it is reading true, and something in that jar has not yet decided what it is.
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "I'm picking up the threads from the previous addendum—QUICK as a non-band, the drifting spring problem, the Bound floor—and now focusing on the retired foundry hand's jar collection as a secret archive. I want to introduce new SVG artifacts that feel discovered, develop the mechanical and conceptual ideas further, and leave open threads for the next section.\n\nThe foundry hand's jars become a chronological \"Drift Almanac\"—a shelf diagram spanning decades showing continuous position drift at Site 19 over time, plotted on the ternary diagram to reveal cyclical or trending movement. For QUICK, I'm proposing it reads things still in their becoming phase, which opens up a new instrument design called \"the Loom\" or \"the Quickening Gauge\"—something that tracks rate of change directly rather than settling position, a fundamentally different mechanism. And for the Bound floor, I'm connecting it elegantly to Site 19's pre-gap position: the floor's numerical value matches the drift vector magnitude, tying together the asymmetric readings and creating a resonant motif across the whole system.\n\nNow I'm structuring Addendum G with four main sections: the Foundry Hand's shelf revealing decades of drift, the Quickening Gauge as a new instrument design, the numeric echo between Bound floor and Site 19 drift, and a teaser about an even older unlabeled jar that opens threads for Addendum H. I'll need to create SVG diagrams for the shelf, the new gauge schematic, and an overlay showing the numeric correspondence, then write it all in the same archival tone with courier notes and open threads at the end. The Wax Ceiling concept ties together the extremes—Named-Twice as positive saturation and Bound floor as negative saturation, both expressions of the same material limit. Now I'm expanding the foundry hand's archive to reveal decades of private tracking predating the official lean years, suggesting he was running his own long experiment before anyone else noticed the drift. I'll include a transcribed excerpt from his ledger in his own hand with cryptic entries, then open threads about whether the Wax Ceiling is a material property or a conceptual one, and whether Site19's drift follows a cycle with implications for the Survey's calibration itself.\n\nI'm also introducing the Quickening Gauge as a test of rate-based measurement on something other than a newborn, and the mysterious unlabeled jar from the foundry hand's pre-ledger period that doesn't match any known spring—possibly the missing fourth vertex or something stranger. Now I'll write the full addendum with consistent voice and styling, keeping the SVGs valid and readable with the established palette.",
    "signature": "EuU6CmMIDxgCKkAmKAtyZ+MfWbKghplwYNS74y2zLXhXza6jBzOcDHHCCjyq3GLciYnYXFQFl1J5pGSOzB2ni2FAc+6/LR7xzcSOMg9jbGF1ZGUtc29ubmV0LTU4AEIIdGhpbmtpbmcSDPfi38Ivi9KUONNBBRoMAkHttX9WTYi0NGTKIjCaI/aM/opK48AFomwgO7ZPliyyEJFRAOdAoagrUTtyP0d/JV0tjMoVJI+T9CdjA/0qrzk+2obo1QQH28uNAHcvq10sLobVGf/YcjtUwavPAuAbO1vTI//zdcOf6iLrJ5K07Zq1qi9QelMNEPyON3yUdwg+AfUkKXuX4hy2zFJiQN76XAPzGGqMTmbVb5Qyyr4+nACB8bezfld6Hj1gADlSvvNhGQ4ltGvvKp4HUGLX/+nNhCFq143SH8VSi2BKH/WciAyo3fvTJT9Aeo09QnEnVrtJaS9PTS9RTe/eyOSXqirMMOnUUZ1dD0Yuq6ncraeAYg7s5SJCjd2PdScEREPICeYC+cr40e9HBTsVulKgxHdKWZ0HYvqt6lVPEO9o5mYmnZsmZicV9xzQsw7EIy5lL1/LQ2ZKCztAUvalpDHtEdT9NIlkwpwr0lfblQqqkCQEwEY089T4F0lH6vlVRg4iuoOanoWiJPCZHDDnEd/MHe1S/jr80Cp+Ilft3DvzH0tgHprxJ522q/55dw/Rf22j83yMB4VQU9OvuNt5Mo/4jr/m5kI7e9YnH9qAhKe0erYCft53OoTKffXZLLfYhnEM4YUhKtLBLh485hDNTed7QKfjIei32faoMrjhB6s6B67ndK9danoIpVjF+jPHyGsDvV/1mjgN+FglzP3GwOySvG0Y2WBjjMHVOECH9a2CZPF8LfOHhI3opCQZS9zHfpnfaN9IrO4b46chPxw10+34qTyEWqlfP189alf7FJdWCP/CUQLlXbge5/75DKpJ+EI752vGPWj/dH7BQD0jNDcy+rnIda/FVTwogKK3fqv01bHXolukh0QpmjmL3CR7QMoFv0gqxupS+CpkRBBlCI619GB/ZUUUFIa3vuoe9x+QuOwRqay6314GEVcLouiySR0CZyzDQkF657ysgzgNjAHprZrED+XpKx7O6T6AQXGW3KW9pNU0BQKFPWwsH0ebGpOTxqjuOMXW+j2jP/u5/kIWcLhpxx1Q1rRXMwUBGOcnfFeiaXrLCqSE4SZNPjZfVQxNrQhAngg2l8r7815f2T78m+9bqO8CWCiNf5lyKd8Lft/YxG3CuphObxb9MMagyA0gnudmAKfL8luag2Qldk0JysaO8ApHkTxPl5047/z2gdT5wMgM5Ttu7Mx1V1PZSdQvLE91s6C7/TIrIddJwy3WmTyr6Lh7To26JKHAOXBJQX4Vw/tld5N7G5a3ZkyWm0PTjqI7dz2hS17CQ2YO+5onv+t4J2MpIv1YhG0OB4wARkIKlilibQXjeb9shgmxvMV3Q12T8Iw0JRRYq+NcIeRSpZgGbSiEtmDcCG9S8E5chYyuHLXohmfDnO6rR6pGvycnq+LjBThTMNH9bHjfwY4oxbybIgwDWY2RlrbPhGUR6aWQMrTBalOA64qiPU+ckOx9x10d6NFPxdwDPBDcZqonr9DkAhuWXmeMb+IYyiuT2QG2Bw/ezMTQqfF6AT2BwPEtOtx21+ML7suzLvSyHXMdVclZVdY4TV+XQP4VqVu/VB/aYYj0S1gc8wlRd6Y72O1Ztb6wSoumRRKOTrWz6/RXu8g+hMZlIprjPjcxv9Z5xbkEkaAhgysN5gKWYkCnOjOY3sNOiAg+0qb+0EociDz3LzX/OCSgvjHYVnCCiiobNEIWxK3NRabaqoR2auy5wrnUFAg/2BI0DhezMbe4ssg2pukUACh1l3JoimzHJ6vWEA1Z/jhuDRV8A7jFmVozvnfckyY6qe4SXJy1VUHD4wBxBgnGKdlSF7MUXVFr/JB8k5ZKQnk4ycoHiOwS4FoKSE74l8RHbEeXSsE7iaY3NGD1C5U4OEpRVWxLQpTBbQ4IqQy/SjqA5aMcRMqMwYVEI6qOAilCrw7stTBKxKFUlLCSuTE9tsjLbHw+D99nLFB6qEEAjQZi2VIG3VZgvxmC7cTjdoO12Hrut8xXwjlNJGvR07+0OcVGmcYwN/gUTk5wG6njfgQb+uX59S39nZ3aMpS1oYBTBOWZ1qnRCOqtVkrPlii9Hu4L0rf3G7EoDg1dwTbzsAbLcGZghOEa/AZ/xkSHQOkNkNC3Ft4l4Si/fj8GaGB6CvuxUAe3P3XMc5MzkvHYMCzr2c4eeTq4ukN4MLr/dU0y6DI18NYlh+riZ6QOktSAXQlncT6Z2T7m/tdNhN63TPOjAKUuQ7vkAe8JdAKXowYLOCTmE3BZRNgGBJsHdDhRY1D+r2znbAP6HENF0rGObTLq92xy+OHp+jORNW5OWLw8VZ4CpZP1TSCGuMbw27dqfNOKTQJ9G7iKcDk9aGB0CiuObgQlMh8LlQ8U+M2/LHJ1D/uSJ20dyJjlHgyG0dsoI4GPP8tOM3s8pJMaoT7HFgZs1UJHWAUeQZkgtIwiDDMjxgpTvu0+StjZAj4NGeuTW9ukfF5yW6AvCYjABm/qsKtZ1UofaaV/FPwMncmdawB9pXZfyfbcsq4hNCiSo15a8pxMfQPpdVgiH2sB7gInG9hmDRbAIUosLInceUPOn40b6W0GVBJR4jXdmjabKO3Hc7BSzk2esJBdr2GWg8GurERJE+5Gyrb9f/vKuJQp9NCtWWJNwdI7x3vzmyq6zYb1YkkOwSQ9v6kU4qcmlUriY5U4qkGL38wyh8QMFOhojXbk1bu+gH7DpvpUii1ojpUoxrHg58uzFib5N0HHKrHQ+t9YhAIOuj7XppB8YCF/AEeGvU80tMvbzMAhBoGYFKAxPkq3VCsFRCbXaEbPy5PSA6lI/+4hDGcTXwME/ZdpE2fEkKseM03RHp/31K3gfC4C0XNNVs0K2tBbXAgjchPSYdDXzQh+CmqKBAzkDkqDx+f98eD/he/R3ms32pgw7GCrsKh6Skl5pzBFvG7xkPDeXf+wb2yIwcjgC89msVXsQs1JFERfGwMHz/xkQub4+bYWUA1/00cf0FONc1qSVQ+lUB1qPmFlZxBvaHPbMmmiLUd+/G5/2nA4P32x5xwzxu7e+OO3CX9lZhFZ3oOK5QomKrdQOBVUd2bMKKbE/+EM6vTCK8DwbTJ4otSV1n2sMfBd0IEu5pJ4p2Xk0Z7rbse9JAKhE1ZyM/YN+rs7exOinIr+NXfuMfQhsBvh85JqZWsTESDpA4YqO5eynbR1IrVicJRyHU9jZAVCIK0Gt+CYbFYWUb7w743y2PSq3Zz6vWC6i3gGXC5H3x64Uv/Xvi1PslEzyTpUzXpKhcauSPvb614StCY3lAErUUsXxwThEVZRVlMV2PjTDzizyN7EIB8uLedXTYaI8kpEq4obvnjtbExnxDiiOSDg1jlaU2HyxeXbgVAKk9rZsc1uO+kUIKQFHSaWn2EyCDcFWG8gWCVPdyYBgvNzc6f1bTavATDn7FD6ed0WYOofHKXN+RAS7dApaIcXTM9CDQ5WkD48aEV3uK/lhLg6T8eGPmIT6+6J0tC/NwHZqCzXZAp3VI2gNY+OUuJO+qPYjZqARnj0IJ+xNzr0ggUrj+t14XNYo0Soy++m0xTlP01w6SwuV8cavobLHuDOiwp9Vie2TzgsqKuajgLRuR4b0yJKq518ICGPXRmja5jI/+lBksmetRUbxA+Bwo9gQiuAUfdZGe0NoKsiZPDwLhgf/2RZLKK56Lz150IGVX5htrzupwhTqdtfdc7Ce4+O4pfv71UAeHuJ/s/yUX66wDqAem5/2CIEMKTjoY9sSqQ9emVejbIm/25F2FQ3nRxs3VhHI3C5UKuH58DtpBrSM1Z2PVxIzTcMiK4JVPYOmjL3hOH+Cs9V5C+aJDScp2WE/FeEac9fpMpdAP5olO8HYhm9yhCMjbkbqV6foVBMB+RGnI20srd0VkYChD9XQGYgmtPjn2pWJFSXXrSVB6fygpqe3ESwPwjCPwCUE72fDLvoikW62OTwMcVplH1F+iKFvNAHw0zB8vMACcJkMm/IPu4b0ogu7rxZlPezOjj40qL3z2RvpJ5OFps47hU9+BXW2iQ5ghdA3BMj9qHDduZ2ScMSAuMCzt8QBw2pk+l9Kgvm3yOuRG3VQXyHlbIML2jUwll8vUgC2inE70LrFzmiRrplacQDpq2MuBYhEzwTAO0PUA102g60UdHUTlF/hO8jZfQ05ukojRHmEfMAsc0NOeTxaU+s5z2WpAuhX22TQSDRrVFL/OVcGQ2c8TvCOqRLo1OGMVe7jtld8+ueXTk0dTZiig0opV+HdLG1QO7MPGDRll7Ud4qUT+7IsEZfgrMQliJZzMITZnEYcw+aznqNuLhjTFP0CVYKIqOBcKb1nMCutwsg0QS8sgIywFmUjZkZxo6RClx/qNNPuy/OUkYdudWQfHQIg+eD/Gdo2t7Hi8Luo+y8V1xW/ZDgSFYxlLA+J3FzJvpJLUAmI6Ji7i5MO6PifBOJRBbo1TzAbkuaHv6L7wRQecKdtQVxCLTDSOfgFc4XHTyKb/Q+3+jAE74do7FVTojUY6WAOzS/rD3+ArOOYRF25HJi4b++RlPU3U2ORG9TML4yYsnxnLZyIghVtrDm1+JC0KlcXPtbrNDu5398Kl5zc0Plw/G7SrejK4S/8gCKlXLFbGUbWvef6Q2u4IoNssApWCoFYAdKYl+gdzlVikm/kHMG3r9FRxHpsXenripWuAqE5EdBvTCke78Jx6v1+Sb/BomLEKAMO5RaY5WaDBP5zK3U6ZqeW2Y81UP/3CiWIpX1BUN/AA6hYrvxDBPFJmtGO8c8K251Rsp1daFmp6RUYWr4YJPua9pks3QWQv2cByLBXJD08LyiXBx+FFgWjy61Di9JlLsY7PLTGETxBJm433PqvCgaVJuHlJ1MwP4QvNMY3X16eEU8E2Muy7RauYNluEdfvdDJSXvi+3GghuS7SYyT4DOoLu6GFuLME+UJbBFMJ4GeLsBi0fXpbtqQq1pc4jgRfF21NF+BO7OixG04tKNhQMSEfMTqeFsnEEGkzfG6LtcgYWRXw8cUCN7TJYPvCXNFpVDIu3qP8FrwYJZnzYvOwurWYI8LG5LKTVmkc8uL48GMZJiAPQdG2kWqRQb7EqCb5cIb3CY70gmxwQ15Xndqp5Bs/HZgD60q/XFpoYtA4JscXnsiMXJddZWIURg0FVk1SvEi7k0KXFC1ybDBnxfAC34IxyjtbKV7kLcEIAJOL73TXHlRWWPZMjCas4ajxYLRHOziFV31P2XPXkgg3EFoqmDkTdDSWxxc3SdhPmq1trWKrYxnWpdHoEvooz86hB3vMYsFSOGDNi7EUKouK4IA2rZaatRDs0Ho4m+9EmiUrTDSVQXo0pIA8LlgHvjUHn/CIXggmHU+81gx2n5nXD+ouiz3TLuMyjr0ceMlf187zQI7C1WyiDXN/I0JkR2p078590ThZZAeAgORGE5RmNOzf3v0HcFWAKsptuC5KK1U+TvddHHc3GLk1ECqf639EsUbhP6i+i4+I5ylPhduFfv/8KmWzaYWM75drx0oeMoMII4fsqAUQ6HXTlMaNC5stg05MyLWJFwo95G/1orVz93F6wSZ+sc55+Ctd29sOiYiVO8so1eeRXq1J0uaBUkKMG/TpPBlJV3o07My7tsPeHTvyaeUxO6yST4cUIX/C4+4V76ZM4L4LRNitnnsXWGIaJrYt9rygVO1yNE8SEEK8Zbmgcd20iw8NUuVqwpLkzbcVSLc24TnKhAgNw1oYMRnr9N4YxKDRNCJppaBKoLZCy2XowtsXMw5bvtnYTBF0Rv5uPkdAaydK0IOLUbnxIfVGup37WHL1Rwx1gMgJRw82JKh1YfTiqt6oFm/TCXXdZQ2QRyS04HsfYftF/nUWBRRwpaSVWM1n6P+w+0R+kAv5HGPRNSwgaqxdn3AftYr9lCHp0+3sbMPIITA/ksCFd4Y3yeNF8Id9FZq/ZMl+MPJQm5fwTHejI2BsF6Jj1AmjItk11ItbSf48TDNptFOGQomnPXABpMQ5PEYsotAWatsqToZ9/syQVunYNILWUpRyv18JiN5fEvYGtTrYAtFP0/CwEC2ZbTxGgCSGiCUk6HhZTD14+p7rqZ8emOFL+hovmGpu6SjKVSUPyelORLCc1lyLjYqMF0A8U8FMIsG0ibSAqKHBfv1xwgP0T34p4qI+5hvUSxgwqd8tSte03/ca3FGfpANy2DD1EuCdkV3aNlcXc08seCGDmQUwyHZ+5/6HJfXRIepAZ/bHA36Z5ytaBjfPgYWWYET+lD/HpjLOdUpV1xk/9Yg5AhNCqGMeMtxp6l9wS8xpj1oo6QOrvXGXLro2OZyRoCgCn0aBJuNekEqyi6/kSRXXz5b67pppmYSjc4H3s38f//pxGCgLBPOnW636dgXpB82CtcqLyuEQ6v94nQS17kd605M9rpsDtDm+NpaObhxHZ8uGnZb7lVXJ6RTXtVofoi3/nlvWjqqjTO+wO6ZxMX7xJNpYHag71AcNrmlYWzrGxe0ePqu3A442205ZFc3BJLuc52258tPUammupZbPhbwF3/XBBrMiS+vw0kv98HG+xgPwtLxpZXrnz9hzNIUMeXXUmE9lny755LLf4hd4/w966AjPXdGz4t6zCmQl6ItUJJmExTrIFk1yWVFaDYWA6m0tLxPZnu20a92XLxGb5LdwAqtrDa8ok5wWC95QRiaJQLmIe8/mn20DfY6iFa6+mZwSx07MiiHmA55bdvS1KPaywz/DfV2uhA+dm65R7KPLHLka+aXeyp6SLRUvNDDiko8PDLnfRVLKsAjGyksnNMgEFdq8lvOtBxZhAkON4ThwUBoy8Kt1X2LGE87A315xssskhXQMyes+T0Y38C5S85/UkvHwfbs7ZHvdDwxcpMBOpQute6hol6E3yskaPGrCgyExRmtTBFuBhA0lEuc5DPbWFrH7hAXeHVWExLmtQ011tmZUcBBOVtrRSrDb25U+66utEluWsSO57X/+F48+ZbMUtyBpKjemxq+7/Q3BuA5F1bttJxrtoIzeJyQrn77gTZOCgYiCK+M6a+/GWzD3GOxdN4l30XB0fH1O5r1+i9gV9BoAZ4+bSbK4z8hMkAzWxegXdwbJj6kVJN2+Kvblw6pNcEGs5SXs3EUErpu8Gg/AwGhMUMeBhZ+5mdd6jezXVMH9+Qq/cqKON14bAIlPrE89pwzPQQf2+WGk82ImMJZXe1Ah9aT2C/0gqGxK4OJpEeRswnGor1U8pPxGZqWlbefjtdTpVQ/YUuKBmXDBzoXlJE5AhRl4U7UntCmeoma0oMz1764JUA5VwmC33tHe5HR2zhupEMBTh/8chmXK/MN7n3+CuzBqoY6E9PFkJOnRUv7rTRp4dgZh38bR3XTsI0ReSbxyH6XUmOmDNE2tE8yedE6cME4mbtiKrw+PX0Cq7ywQI8Vk8GqDLh8D0yi5Vgwjo0BsalDgSwyWZmRikmyIYEf4cbzLlouco/a6oaD3ubqf+4RZ9/DDuoEfDjlhx5dJsLaB3ej6nnsO4Fx1NVDnbPW7Q3hC8s6k4ZLeQA5wAebwzaKnOgbkCupqvSEbrQQGnqSyH5l6cnJJtN1kluZfNr9gveTWg3nIy6+/NxludQR55/tDzRtukuU9LWxWsM4LpHUXGvR8LXZacWXdNzdga9NDPIjOCDDJpF6wPNtHUpUChEsLg9pJoqfopox3dXEdvkcTDLufBo9tvw44bQdxTRR7x6riKCwqPFOii+9b11O2Y2vIkgc7xF1fFVUAwASFJ96WFSOIwgq7DmD7n9KZkhhQOawLxPoZOsXkp39FbJ29YtWguIxqAxOU8BZyYfO60BkWgPbWImwwaKKVUPyJiq/MvCR8mQfUnCZ/Mbjeh7wUioI+LRKFJ0XUSEElofaKDwq2Q5iRnNhBSx2Qy9OPpvaOPRNfd322NmzbFnF5KN18ytEDT6+IBrED8stK23J7Ee8FG3cERiHymrNEPa/rUfHKEBN2OMHr31By0dg300bMQ0JP0qyYPDfAMvFoFF1YP4mMX21x6FZUOrhClBDA6/NGXNIFli0I5hjI9Om1JvCfINQ0+X+ryoGh+cC8GCY8/H2c2fAHBOoppOH73P9H/iSDpRZEla6qye5DPE1evhfNm4M+0vmFnHMxMFWphLbXDGxI0Sa1DUr9wovKtKJHNmWJz9lGqqGnoOiNZ5pKk5CL6jE4+rj6NoG4BEm46IFBwEMBD/jsO1tNPK+RydRcojwE1D84mIjU4n383M7o2r/ZLi+CyLOENFb70mU68k/cZo0/kxskx2RwW8PiKiHDiovChy6H4Gj0m40hRqc8iEbFYKPqHMWtWIy/xtDxZOJPysx2SQ56JBnzTSllbfEt3R6Myy1bDhtl3ISqcuq27nwPyS3v7DL7t2zr2xeOKzFjD+bypRtcsREWVWgWScpeGRXky64Nl0YfnbWyX8fdn1ymbFj2kwS34VO5O9MoKR1aE0Rs2vQsYqxeK5SWNGewzU4KQwkzIj22njqbd2vH/d53v0z7t6dd0dNbDAH/gQgsEALi+Wn5zEDTrPsdm6O7lSFCpWzsmLHghPwKIWwpc3eskJw4EX0w22z3v39Zx/wGz613yXuB8HmMWn4pMT/a8ZXS5zNgCDQeUW0VGcFymDuEULaiLam8oY2jpM9uqPtEROayZQuTMbsGSgYzTY7RGhHA+lYk8cbUbXiojRZVmIUH6x8oZW/6g7LNIJdVNlecPeEvYb9JbkhHqpH3V1TTG3C75GvRfJCATgzKGvVjR2r8V/N2cALe1YMUqAXnhRF20kMnfwarB0uwM6e75v7DHx0TSe3tXic2tRtIkKTePy4EqQkQlqf6PP0WK9PpRx5kO4HUdRYkTg7IIXIba87/WKUfdcoentXKWapwsvdZNS4tltOjbanb2mDMB1BxuTRIB+57PteGyKkbQKFAYK0jgthxxG+YzW1idDSrx19qIrkI5bKtdkAh3QqdIrssg/bVNYtiFQyN+TlakluKZs1cpOfqVjFvJTQXZ6nrnKzeMUzDN69bvCWJGweSvMAIK5XngTtxmLFUtOble1DO3/fb+mggg1hPlWxk0ITy8WHDVhGzDzbmTN5YUXE4t6Xa/znDUEKTVXWxl108vvvu1sKhc/0rvwRyH6h5/pGCTPwglGFxwwzzWLjz7fuuH4BVfJaUd2cIimaxZjWVCReIgGKMsVdYYJgDKyzHz5zudr4/mvTbmo3nrD3eX2fFM1xahjLUzcSCLY5dVZzNZqnkVQaHwF2xpVCL3B4wYDV1xALBpkN44IdExRijaQb5O23R9fxn3uVFfFMb+Ikt0w8xWMvDSYU5+ZfaUwOVDjN9e68Cnt5scgH3lnbg6URyeEGXGDnio6zy7zfswYIGDwzlcY1KZd7hi3NBM+q264/8q85x5vh3lPcUHWKODX30bsjAwAw6qzQg2MvrBFQSTxcaEPefmprACEwPNurR+3bSZfmQQGcs44m3PR5U1SIQUsGIXHTJtsZqfJNUX0p9td0CYagjMAFPzLXJvbNdiJbhCzlLmfJpIQuSxd++E20fjHm9DrWe7ZOi4KQT1Xgd5r9VJuliY1InS+8TBg3kWGruJCm9ay8kLW9Tmmr0wnZ3PbjWU3qK0mQPUzI68vcxpvi1ei4kJYsMFDd7Z76UTCQYjvY0DG2quIgu5bvb9jyEMs25vYSt/+8ezXye4TtJBtYY2AvM3BcYYzezLhJzilCGQbWLQDOkDTZLHEteCPCK3S3zNOojROuMOTvV4z+vM3P686ph2CUs1hz6aYrQ4b/JvslkKUAXTUogDgdll8OE7cElPDmwInE187Brf2k3yZYzDu6SFMl7K27MWjPr4NF+TJ8wK6dznlzgIhYjHZH6HtB7pGbZNFNAWzFRIJSRXIqxNc0ghbcoXEbgQzp3Pxln+q4BE5IlmrZEXURVmeJAWJabhMofECHzxgB"
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM G\n### *The Foundry Hand's Shelf, and the Quickening Gauge*\n*(the courier left this one inside a jar; the Office returned the jar, empty, with no comment)*\n\nThe retired foundry hand — let the log finally use his name, since he has stopped objecting: **Coen Ashery**, forty-one years at the Site 19 works, eleven of them after retirement spent doing something nobody asked him to do. He had more jars. He had a great many more jars. And somewhere in sorting them, the Bound floor and the Newborn fan turned out to be reaching for the same wall from opposite sides.\n\n---\n\n## I. The Almanac\n\nAshery's shelf, when finally inventoried, held thirty-eight sealed sample-jars of raw Site 19 casting wax, dated continuously back well before the lean years anyone else had noticed — the earliest jar predates the foundry ledger's own first legible entry by nine years. He kept one jar roughly every fourteen months, unprompted, \"because the water never poured the same way twice and somebody should write that down even if it isn't anybody's job.\"\n\nRun through the Kinship Assay against the same three references, the jars do not show a single drift event. They show a **cycle**.\n\n```svg\n<svg viewBox=\"0 0 480 300\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"10\" fill=\"#3a2e1f\">\n  <rect width=\"480\" height=\"300\" fill=\"#f1e9d8\"/>\n\n  <!-- shelf of jars, schematic -->\n  <line x1=\"20\" y1=\"60\" x2=\"460\" y2=\"60\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <g>\n    <!-- 12 representative jars along the shelf -->\n    <circle cx=\"40\"  cy=\"48\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n    <circle cx=\"78\"  cy=\"44\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n    <circle cx=\"116\" cy=\"40\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n    <circle cx=\"154\" cy=\"46\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#7a1f1f\" stroke-width=\"2\"/>\n    <circle cx=\"192\" cy=\"52\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#7a1f1f\" stroke-width=\"2\"/>\n    <circle cx=\"230\" cy=\"48\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n    <circle cx=\"268\" cy=\"42\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n    <circle cx=\"306\" cy=\"44\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#7a1f1f\" stroke-width=\"2\"/>\n    <circle cx=\"344\" cy=\"50\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#7a1f1f\" stroke-width=\"2\"/>\n    <circle cx=\"382\" cy=\"46\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n    <circle cx=\"420\" cy=\"40\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n    <circle cx=\"450\" cy=\"48\" r=\"9\" fill=\"#e3d7bd\" stroke=\"#3a2e1f\"/>\n  </g>\n  <text x=\"240\" y=\"20\" text-anchor=\"middle\" font-weight=\"bold\">Ashery's shelf, 38 jars, ~14 months apart — red rims mark \"lean\" readings</text>\n\n  <!-- the drift plot beneath, % Site-19-of-itself over time, oscillating -->\n  <text x=\"20\" y=\"100\" font-weight=\"bold\">% match to MODERN Site 19, by jar date</text>\n  <line x1=\"40\" y1=\"280\" x2=\"440\" y2=\"280\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n  <line x1=\"40\" y1=\"120\" x2=\"40\" y2=\"280\" stroke=\"#3a2e1f\" stroke-width=\"1.5\"/>\n  <text x=\"20\" y=\"125\" font-size=\"9\">100%</text>\n  <text x=\"20\" y=\"282\" font-size=\"9\">75%</text>\n\n  <path d=\"M40 150 C70 130, 100 200, 130 230 C160 255, 190 250, 220 200\n           C250 150, 280 135, 310 200 C340 250, 370 245, 400 190 C420 160, 440 150, 440 150\"\n        fill=\"none\" stroke=\"#7a1f1f\" stroke-width=\"2\"/>\n\n  <text x=\"130\" y=\"295\" font-size=\"9\" text-anchor=\"middle\">lean trough, yr ~14</text>\n  <text x=\"350\" y=\"295\" font-size=\"9\" text-anchor=\"middle\">second trough, yr ~33</text>\n  <text x=\"240\" y=\"135\" font-size=\"9\" text-anchor=\"middle\" fill=\"#5c3b1e\">period ≈ 19 years, not yet closed twice</text>\n</svg>\n```\n\nThe \"lean years\" were not an anomaly. They were the trough of a period Ashery's own shelf catches twice, roughly nineteen years apart, and may be catching a third time now, partway down the second descent. The Third Cataloguing calibrated true Still against Site 19 at a point on this curve nobody marked as a point on a curve, because nobody suspected there was a curve to be caught at a phase of. Every reading the Survey has ever taken is phase-locked to wherever Site 19 happened to be sitting on a cycle whose existence postdates the instrument that depends on it.\n\nAshery's note, copied from the inside lid of the newest jar, lid-side, where he apparently expected nobody would look until after he'd stopped being askable about it:\n\n> *\"I am not smart enough to know if it breathes or just wobbles on its bedrock the way anything heavy does over enough years. I only know it does not hold still, and a spring that does not hold still has no business being the thing that tells everyone else whether they do.\"*\n\n---\n\n## II. The Quickening Gauge\n\nAddendum F left QUICK off the scale entirely — a fan, not a point, eleven minutes of motion that never arrived. The standard instrument cannot read it because the standard instrument was built to report **where**, and the newborn's needle was reporting **how fast**, which is a different question the old housing has no dial for.\n\nA new instrument, built rather than adapted: not a needle over a fixed dial, but a stylus over a slow-turning drum, recording the *spread* of the fan continuously instead of forcing it to a single settled point.\n\n```svg\n<svg viewBox=\"0 0 460 280\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect width=\"460\" height=\"280\" fill=\"#f1e9d8\"/>\n\n  <!-- drum, side view -->\n  <g transform=\"translate(70,60)\">\n    <ellipse cx=\"80\" cy=\"0\" rx=\"14\" ry=\"70\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <ellipse cx=\"200\" cy=\"0\" rx=\"14\" ry=\"70\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <line x1=\"80\" y1=\"-70\" x2=\"200\" y2=\"-70\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <line x1=\"80\" y1=\"70\" x2=\"200\" y2=\"70\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <!-- traced band, irregular width, recording spread over time -->\n    <path d=\"M80 -10 Q120 -45 140 -20 Q160 5 180 -25 Q195 -40 200 -15\"\n          fill=\"none\" stroke=\"#7a1f1f\" stroke-width=\"1.5\"/>\n    <path d=\"M80 10 Q120 50 140 25 Q160 -5 180 30 Q195 45 200 20\"\n          fill=\"none\" stroke=\"#7a1f1f\" stroke-width=\"1.5\"/>\n    <text x=\"140\" y=\"-95\" text-anchor=\"middle\" font-size=\"10\">drum turns at fixed rate — band-width = QUICK reading</text>\n  </g>\n\n  <!-- pivot/stylus arm -->\n  <g transform=\"translate(140,60)\">\n    <line x1=\"0\" y1=\"0\" x2=\"0\" y2=\"-30\" stroke=\"#7a1f1f\" stroke-width=\"2\"/>\n    <circle cx=\"0\" cy=\"-30\" r=\"3\" fill=\"#7a1f1f\"/>\n    <text x=\"0\" y=\"-40\" text-anchor=\"middle\" font-size=\"9\">stylus, free to swing — never forced to settle</text>\n  </g>\n\n  <text x=\"230\" y=\"200\" text-anchor=\"middle\" font-size=\"11\" font-weight=\"bold\">\n    Quicken Index = band-width, in degrees-of-arc per minute of recording\n  </text>\n  <text x=\"230\" y=\"220\" text-anchor=\"middle\" font-size=\"10\">\n    a narrowing band over successive turns is read as \"settling\" —\n  </text>\n  <text x=\"230\" y=\"236\" text-anchor=\"middle\" font-size=\"10\">\n    the gauge's first instrument capable of detecting that an object is in the act of arriving.\n  </text>\n  <text x=\"230\" y=\"258\" text-anchor=\"middle\" font-size=\"9\" fill=\"#7a1f1f\" font-style=\"italic\">\n    R's niece, re-run on the Quickening Gauge at three months: band narrowing, not yet closed.\n  </text>\n</svg>\n```\n\nR's niece was the first and so far only trial subject, gauged again at three months with the family's permission, no Name anchor still required since the instrument doesn't ask for a settled point. The band has narrowed by roughly a third since the naming-eve session. It has not closed. The working guess, unconfirmed, is that it closes somewhere near the time a child would ordinarily be expected to answer to a name on her own, without it being offered to her — which would mean QUICK isn't a Band so much as a **countdown to the moment Name becomes available as an anchor at all.**\n\n---\n\n## III. The Wax Ceiling\n\nThe coincidence Addendum F left lying in plain sight: the deepest Bound readings across every Yoke bead trial cluster around **−2.3 to −2.4 Cauls** and refuse to go lower no matter how many people are melted into the bead. Ashery's pre-gap jar, run for the first time against *itself* — that is, against a contemporary Site 19 sample, the closest thing to gauging a spring's drift directly against its own younger water — returns a number from the same neighborhood.\n\n```svg\n<svg viewBox=\"0 0 440 180\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect width=\"440\" height=\"180\" fill=\"#f1e9d8\"/>\n\n  <line x1=\"40\" y1=\"90\" x2=\"400\" y2=\"90\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"40\" y=\"75\" font-size=\"9\">−3</text>\n  <text x=\"400\" y=\"75\" font-size=\"9\">0</text>\n\n  <!-- ceiling band -->\n  <rect x=\"62\" y=\"60\" width=\"40\" height=\"60\" fill=\"#7a1f1f\" opacity=\"0.15\"/>\n  <text x=\"82\" y=\"50\" text-anchor=\"middle\" font-size=\"9\">the Ceiling</text>\n\n  <!-- Bound floor cluster -->\n  <circle cx=\"75\" cy=\"90\" r=\"5\" fill=\"#7a1f1f\"/>\n  <circle cx=\"80\" cy=\"90\" r=\"5\" fill=\"#7a1f1f\"/>\n  <circle cx=\"84\" cy=\"90\" r=\"5\" fill=\"#7a1f1f\"/>\n  <circle cx=\"78\" cy=\"90\" r=\"5\" fill=\"#7a1f1f\"/>\n  <text x=\"80\" y=\"110\" text-anchor=\"middle\" font-size=\"9\">all Yoke-bead floors</text>\n  <text x=\"80\" y=\"122\" text-anchor=\"middle\" font-size=\"9\">2-way, 3-way alike</text>\n\n  <!-- pre-gap jar vs itself -->\n  <circle cx=\"88\" cy=\"90\" r=\"5\" fill=\"#3a2e1f\"/>\n  <text x=\"88\" y=\"138\" text-anchor=\"middle\" font-size=\"9\">pre-gap jar</text>\n  <text x=\"88\" y=\"150\" text-anchor=\"middle\" font-size=\"9\">vs modern Site 19</text>\n\n  <text x=\"240\" y=\"155\" text-anchor=\"middle\" font-style=\"italic\" font-size=\"10\">\n    Both lines of evidence saturate at the same wax-physical limit, not the same conceptual one.\n  </text>\n</svg>\n```\n\nThe working theory, stated plainly for the first time: **Bound has no floor of its own.** What the gauge reports as \"deepest possible commitment\" may simply be the deepest strain the anchor-wax can hold before its molecular memory of position stops discriminating — the same physical ceiling, approached from the other direction, that fuses Named-Twice's bob solid past +3. Vacant and the floor of Bound may be the *same wall*, read from opposite faces, and the instrument has never been built that could tell the difference between \"this relation is functionally infinite\" and \"this relation is merely past where my wax can still keep score.\"\n\nIf true, every Bound reading this log has ever logged as a number was never a number. It was a **flag**: *deeper than I can say.* The three veteran surveyors' declined recasts, M and R's bead, the three-way crew — none of them may differ from each other at all. The gauge has been reporting a ceiling and calling it a measurement.\n\n---\n\n## IV. The Jar With No Date\n\nAmong Ashery's thirty-eight, one jar carries no label, no date, and — unlike every other jar on the shelf — no trace of Site 19 wax chemistry whatsoever. Run against all three known references and against Ashery's own oldest dated Site 19 sample, it matches **none of them**, including itself: two assays of the same jar's contents, taken eight days apart by the same operator with full anchors met, returned different ternary positions.\n\nAshery's only comment, given to the surveyor who found it and transcribed faithfully: *\"That one I didn't pour. I found it already on the shelf when I started, in '04, older lid-wax than anything we were casting that decade. I kept meaning to ask the old foundry master where it came from. He died before I worked up the nerve, and I have been too much of a coward to open it again since the second reading came back wrong twice in a row.\"*\n\n---\n\n## Open Threads (for Addendum H)\n\n- **The cycle's third trough.** If the ~19-year period holds, the Survey is partway down a second descent right now, with no record of anyone having calibrated *anything* against the previous trough — meaning the current \"true Still\" may be sliding, slowly, under every reading taken this decade, undetected because nobody had a second jar to check it against until Ashery.\n- **The Ceiling, tested properly.** Does the wax-physical saturation limit move if the anchor-wax is cast from a different spring entirely — Coast, Bench, even the unnamed fourth bearing from the Assay margins? If the Ceiling is material rather than relational, a non-Site-19 Yoke bead should saturate at a different number. Nobody has built one.\n- **The Quickening Gauge, on adults.** Tried only on an infant so far. Does a freshly sworn oath, a newly dug well, a marriage an hour old, show the same narrowing band — meaning QUICK isn't reserved for the literally newborn, but for anything that hasn't finished arriving yet, on any scale?\n- **The undated jar.** Two assays, same contents, two different answers. Either the assay failed, the jar's contents are actively changing between readings, or — least comfortable, most consistent with everything else this season has found — it is reading true, and something in that jar has not yet decided what it is."
  }
]
#11 · Assistant
2026-06-30 21:59:24
Model · claude-sonnet-5
Thinking
I'm picking up where the series left off—building Addendum H by weaving together all the established machinery: the four anchors, Witness Coin, Yoke bead, Kinship Assay, Ceiling theory, Quickening Gauge, the 19-year cycle, and the undated jar. The open threads I want to develop are the cycle's third trough with its calibration drift and testing the Ceiling against new conditions.

The strongest approach is to tie these threads together through three key experiments: testing the Ceiling using a non-Site-19 Yoke bead to prove the Ceiling is material rather than relational, applying the Quickening Gauge to adult oaths and marriages to show how old vows can become unsettled again, and running a third assay on the undated jar with two simultaneous witnesses to discover whether the jar's instability comes from itself or from the observer—that it might actually be a mirror-instrument reflecting back the Witness's own Drift rather than having any intrinsic substance of its own. That last insight recontextualizes everything and ties back elegantly to the Witness mechanics established earlier. Structuring the addendum with three core findings: the Coast Sill ceiling test showing a different magnitude than Site-19 (proving Ceiling is material-dependent), the Mirror Jar correlation study where different witnesses get different readings tied to their own Drift, and the Quickening Gauge on an oath renewal revealing bands can re-open when circumstances shift. I'll format it as a courier note with sections, SVG diagrams, and transcripts, keeping the same voice and structure as the previous addenda, then flag open threads for the next one. rows the witness's identity-signature instead of using it as a reference point, the way an unwitnessed gauge borrows birth-water. It's an inversion: the Witness becomes the content itself, not just the frame. I'll create a new glyph for this — "the Mirror" or "Unkeyed" wax — represented as an empty circle that reflects whatever's anchored to it.

For the Quickening Gauge, I want to test it on renewal situations like a couple renewing vows or a re-dug well. The band should initially be narrow since these are presumed Bound and settled, but if it reopens during recording, that suggests the renewal caused a re-quickening — meaning Bound isn't actually terminal. Better yet: run the Quickening Gauge on M and R's Yoke bead itself rather than a single target. If it's still narrowing very slightly despite the stable Bound reading from before, that implies no Bound state is truly final — they're all mid-quicken at rates too slow for the standard gauge to catch, and the standard gauge just calls it "arrived" because it stops looking too soon.

The Quickening Gauge with continuous drum recording reveals slow residual movement everywhere, even in supposedly Still targets, which completely reframes how the instrument works. I should also keep testing Still-band targets like Site 19 itself or doors on the Quickening Gauge to see if "Still" is actually just "Quick, but slow enough to fool the standard dial."

Now I'm planning the SVG diagrams and structure for the full addendum — a two-ceiling number line showing the Site-19 cluster versus the Coast-Sill cluster, a correlation chart for the Mirror Jar readings across different witnesses, the new glyph for Unkeyed wax, and a drum diagram showing the residual quickening in the Yoke bead. I'll match the style and depth of the previous addenda while keeping it concise.
# FIELD INSTRUMENT NO. 1 — ADDENDUM H ### The Two Ceilings, and the Mirror Jar (the courier brought no envelope and no jar this time — only said "they're going to want this one back faster than usual" and left before anyone could ask who "they" was) Three threads from Addendum G demanded the same kind of answer: stop theorizing about the Ceiling and build against it; stop assuming the undated jar is broken and ask what it's broken at; and take the Quickening Gauge — built for a newborn — and point it at something the Survey had already filed as finished. --- ## I. The Second Ceiling If the Bound floor is a property of the wax rather than the relation, a Yoke bead cast from a different spring's anchor-wax should saturate at a different number. Two surveyors with Coast-Sill-provenance Coins — stamped Site 4 at their naming, confirmed against the foundry ledger, no Site 19 in their casting at all — agreed to a bead.
Site-19-wax Yoke beads (all trials to date):           Coast-Sill-wax Yoke bead (new):
   M+R           −2.3 Cauls                               −3.6 Cauls
   A+B           −2.1 Cauls                               (single trial; second pair
   A+C           −1.9 Cauls                                being sought for confirmation)
   B+C           −2.4 Cauls
   A+B+C         −2.3 Cauls
   ── all cluster within 0.5 of each other ──          ── sits a full Caul-and-change deeper ──
```
−4 0 Site-19-wax ceiling ≈ −2.3 Cauls Coast-Sill-wax ceiling ≈ −3.6 Cauls (n=1) Two ceilings, two materials. The wall moves with the wax. It is not a fact about Bound.
``` This settles the question the way Addendum G hoped it wouldn't have to: the Ceiling is material, not conceptual. Site 19's wax simply runs out of room to record commitment sooner than Coast Sill's does. Every "deepest possible Bound" this log has ever logged was, more precisely, deepest possible Bound expressible in whatever wax happened to be in the housing that day. The Survey has been measuring devotion with a ruler that runs out before some of the things it's measuring do. A second-order worry, not yet tested: if the Ceiling is material, does the positive ceiling — the one Named-Twice hit, fusing solid past +3 — also move with provenance? Nobody has been willing to deliberately reproduce a Named-Twice fusion to find out, for reasons that should not need restating. --- ## II. The Mirror Jar Two assays of the undated jar, eight days apart, same operator, returned different ternary positions. The obvious next question — obvious in hindsight, expensive to arrange — was whether the jar was changing between readings, or the operator was, and the jar was simply better at showing it than anything else in the foundry. Two different Witnesses gauged the jar on the same morning, an hour apart, each logging their own Coin immediately beforehand.
WITNESS    COIN (that morning)      JAR READING (ternary, vs. Site19/Coast/Bench)
Surveyor P   0.6 Tide                58% / 31% / 11%
Surveyor Q  −1.8 Bound                22% / 9%  / 69%   (toward Bench — Q's own naming-Coin
                                                          is Bench-Sill-cast)
``` Surveyor Q's casting-spring is Bench Sill. The jar, witnessed by Q, leaned hard toward Bench. Witnessed an hour earlier by P — whose own Coin carries no strong kinship to any single spring — the jar gave a comparatively even spread. Two more pairings since have held the same shape: the jar's reading correlates with the Witness's own provenance, not with anything in the jar. This is not the ordinary Kinship Assay bias from Addendum C. An ordinary unwitnessed gauge defaults to its own birth-water in the absence of a fourth anchor. This jar is witnessed, properly, every time — and still returns the Witness's signature instead of a stable account of itself. The working account: the jar was never cast from any spring at all. It has no birth-water to default to, kinship-wise, which the Assay already confirmed (it matches none of the three references). Lacking a self to compare against, and given a Witness who does have one, it does not anchor the reading. It borrows the Witness's.
JAR no birth-water; empty kinship slot Witness's own kinship flows IN, fills the slot, is read back as the jar's. Witness
``` Proposed glyph, for a category the table has never needed before — not a Band, not QUICK's unbanded becoming, but a vacancy that takes the shape of whoever fills it:
UNKEYED reads the reader
``` If Unkeyed wax exists, two uncomfortable possibilities follow. First: an Unkeyed gauge, mistaken for a clean one, would never show a stable target reading at all — it would show a rotating cast of whoever last held it, filed as "noisy data" rather than recognized as a different kind of instrument entirely. Second, worse: nobody knows whether Ashery's jar is a freak of casting, or whether most wax has a faint trace of this property and Site 19's enormous share of the foundry's output has simply been swamping it with birth-water bias strong enough to hide it until now. --- ## III. The Quickening Gauge, Pointed at Something Already Filed as Finished The Quickening Gauge was built for a newborn. Nobody had tried it on anything the Survey already considered settled — until someone thought to run it on the M+R Yoke bead from Addendum E, the one that has sat at a stable −2.3 Cauls, Bound, since the season it was cast, and which both M and R asked never to have re-read aloud. The standard gauge confirmed the same number it always gives this bead. The Quickening Gauge, recording continuously instead of stopping at "settled enough," was left running for forty minutes — far longer than any prior trial, because nobody expected to need that long for something already filed as finished.
40-minute recording — M+R bead, "Bound, settled" per standard gauge Residual Quicken Index: 0.04°/min — nonzero far too slow for the standard gauge's settling window to register as anything but "arrived."
``` The band is not closed. It is narrower, by orders of magnitude, than the newborn's eleven-minute fan — but it is still narrowing, almost imperceptibly, forty minutes in. The standard gauge calls this Bound and stops looking, because Bound is what "narrow enough, long enough" has always meant to an instrument built to give a surveyor a number before lunch. The uncomfortable implication, stated as plainly as Addendum G stated the Ceiling's: the standard gauge may never have measured anything as actually arrived. It measures things slow enough to fool its own patience. Three veteran surveyors' Bound Coins, the Yoke beads, possibly Site 19 itself sitting at the bottom of its own nineteen-year trough being read as "true Still" — every one of them may simply be Quickening too slowly for the old housing's settling time to catch the motion. Nothing has been shown to stop. Only to slow past the instrument's patience. --- ## Open Threads (for Addendum I) - Is anything actually Still? If the Quickening Gauge finds residual motion in a forty-minute trial of something already filed as finished, what does an hour find? A day? Has the Survey ever, in ninety years, held a target long enough to rule out a Quicken too slow for the standard dial — including Site 19's own original calibration? - The Unkeyed hypothesis, scaled up. Is Ashery's jar a singular freak casting, or a trace property latent in all wax, masked everywhere else by Site 19's overwhelming share of the foundry record? If the latter, every "clean" reading this log has trusted may carry a small Unkeyed component nobody has had reason to isolate. - The positive Ceiling, untested by design. The Coast-Sill Bound-floor experiment has a mirror nobody will run: a deliberately non-Site-19 Named-Twice attempt, to see if the fusion-point also moves with provenance. The Office's verbal position on this is reported as "absolutely not," which the log notes without endorsing. - Q's discomfort. Surveyor Q, on learning the jar had read back Bench Sill through her specifically, asked to see her own naming-Coin record again for the first time in four years. She has not yet said what she found, and nobody has asked twice.
Debug · Raw response blocks JSON
[
  {
    "type": "thinking",
    "thinking": "I'm picking up where the series left off—building Addendum H by weaving together all the established machinery: the four anchors, Witness Coin, Yoke bead, Kinship Assay, Ceiling theory, Quickening Gauge, the 19-year cycle, and the undated jar. The open threads I want to develop are the cycle's third trough with its calibration drift and testing the Ceiling against new conditions.\n\nThe strongest approach is to tie these threads together through three key experiments: testing the Ceiling using a non-Site-19 Yoke bead to prove the Ceiling is material rather than relational, applying the Quickening Gauge to adult oaths and marriages to show how old vows can become unsettled again, and running a third assay on the undated jar with two simultaneous witnesses to discover whether the jar's instability comes from itself or from the observer—that it might actually be a mirror-instrument reflecting back the Witness's own Drift rather than having any intrinsic substance of its own. That last insight recontextualizes everything and ties back elegantly to the Witness mechanics established earlier. Structuring the addendum with three core findings: the Coast Sill ceiling test showing a different magnitude than Site-19 (proving Ceiling is material-dependent), the Mirror Jar correlation study where different witnesses get different readings tied to their own Drift, and the Quickening Gauge on an oath renewal revealing bands can re-open when circumstances shift. I'll format it as a courier note with sections, SVG diagrams, and transcripts, keeping the same voice and structure as the previous addenda, then flag open threads for the next one. rows the witness's identity-signature instead of using it as a reference point, the way an unwitnessed gauge borrows birth-water. It's an inversion: the Witness becomes the content itself, not just the frame. I'll create a new glyph for this — \"the Mirror\" or \"Unkeyed\" wax — represented as an empty circle that reflects whatever's anchored to it.\n\nFor the Quickening Gauge, I want to test it on renewal situations like a couple renewing vows or a re-dug well. The band should initially be narrow since these are presumed Bound and settled, but if it reopens during recording, that suggests the renewal caused a re-quickening — meaning Bound isn't actually terminal. Better yet: run the Quickening Gauge on M and R's Yoke bead itself rather than a single target. If it's still narrowing very slightly despite the stable Bound reading from before, that implies no Bound state is truly final — they're all mid-quicken at rates too slow for the standard gauge to catch, and the standard gauge just calls it \"arrived\" because it stops looking too soon.\n\nThe Quickening Gauge with continuous drum recording reveals slow residual movement everywhere, even in supposedly Still targets, which completely reframes how the instrument works. I should also keep testing Still-band targets like Site 19 itself or doors on the Quickening Gauge to see if \"Still\" is actually just \"Quick, but slow enough to fool the standard dial.\"\n\nNow I'm planning the SVG diagrams and structure for the full addendum — a two-ceiling number line showing the Site-19 cluster versus the Coast-Sill cluster, a correlation chart for the Mirror Jar readings across different witnesses, the new glyph for Unkeyed wax, and a drum diagram showing the residual quickening in the Yoke bead. I'll match the style and depth of the previous addenda while keeping it concise.",
    "signature": "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"
  },
  {
    "type": "text",
    "text": "# FIELD INSTRUMENT NO. 1 — ADDENDUM H\n### *The Two Ceilings, and the Mirror Jar*\n*(the courier brought no envelope and no jar this time — only said \"they're going to want this one back faster than usual\" and left before anyone could ask who \"they\" was)*\n\nThree threads from Addendum G demanded the same kind of answer: stop theorizing about the Ceiling and *build* against it; stop assuming the undated jar is broken and ask what it's broken *at*; and take the Quickening Gauge — built for a newborn — and point it at something the Survey had already filed as finished.\n\n---\n\n## I. The Second Ceiling\n\nIf the Bound floor is a property of the wax rather than the relation, a Yoke bead cast from a different spring's anchor-wax should saturate at a *different* number. Two surveyors with Coast-Sill-provenance Coins — stamped Site 4 at their naming, confirmed against the foundry ledger, no Site 19 in their casting at all — agreed to a bead.\n\n```\nSite-19-wax Yoke beads (all trials to date):           Coast-Sill-wax Yoke bead (new):\n   M+R           −2.3 Cauls                               −3.6 Cauls\n   A+B           −2.1 Cauls                               (single trial; second pair\n   A+C           −1.9 Cauls                                being sought for confirmation)\n   B+C           −2.4 Cauls\n   A+B+C         −2.3 Cauls\n   ── all cluster within 0.5 of each other ──          ── sits a full Caul-and-change deeper ──\n```\n\n```svg\n<svg viewBox=\"0 0 460 180\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"11\" fill=\"#3a2e1f\">\n  <rect width=\"460\" height=\"180\" fill=\"#f1e9d8\"/>\n  <line x1=\"40\" y1=\"100\" x2=\"420\" y2=\"100\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"40\" y=\"85\" font-size=\"9\">−4</text>\n  <text x=\"420\" y=\"85\" font-size=\"9\">0</text>\n\n  <!-- Site 19 ceiling cluster -->\n  <g fill=\"#7a1f1f\">\n    <circle cx=\"285\" cy=\"100\" r=\"5\"/><circle cx=\"291\" cy=\"100\" r=\"5\"/>\n    <circle cx=\"280\" cy=\"100\" r=\"5\"/><circle cx=\"288\" cy=\"100\" r=\"5\"/><circle cx=\"283\" cy=\"100\" r=\"5\"/>\n  </g>\n  <text x=\"285\" y=\"120\" text-anchor=\"middle\" font-size=\"9\">Site-19-wax ceiling</text>\n  <text x=\"285\" y=\"132\" text-anchor=\"middle\" font-size=\"9\">≈ −2.3 Cauls</text>\n\n  <!-- Coast Sill ceiling, deeper -->\n  <circle cx=\"155\" cy=\"100\" r=\"6\" fill=\"#3a2e1f\"/>\n  <text x=\"155\" y=\"120\" text-anchor=\"middle\" font-size=\"9\">Coast-Sill-wax ceiling</text>\n  <text x=\"155\" y=\"132\" text-anchor=\"middle\" font-size=\"9\">≈ −3.6 Cauls (n=1)</text>\n\n  <text x=\"230\" y=\"160\" text-anchor=\"middle\" font-style=\"italic\" font-size=\"10\">\n    Two ceilings, two materials. The wall moves with the wax. It is not a fact about Bound.\n  </text>\n</svg>\n```\n\nThis settles the question the way Addendum G hoped it wouldn't have to: the Ceiling is **material**, not conceptual. Site 19's wax simply runs out of room to record commitment sooner than Coast Sill's does. Every \"deepest possible Bound\" this log has ever logged was, more precisely, *deepest possible Bound expressible in whatever wax happened to be in the housing that day.* The Survey has been measuring devotion with a ruler that runs out before some of the things it's measuring do.\n\nA second-order worry, not yet tested: if the Ceiling is material, does the *positive* ceiling — the one Named-Twice hit, fusing solid past +3 — also move with provenance? Nobody has been willing to deliberately reproduce a Named-Twice fusion to find out, for reasons that should not need restating.\n\n---\n\n## II. The Mirror Jar\n\nTwo assays of the undated jar, eight days apart, same operator, returned different ternary positions. The obvious next question — obvious in hindsight, expensive to arrange — was whether the *jar* was changing between readings, or the *operator* was, and the jar was simply better at showing it than anything else in the foundry.\n\nTwo different Witnesses gauged the jar on the same morning, an hour apart, each logging their own Coin immediately beforehand.\n\n```\nWITNESS    COIN (that morning)      JAR READING (ternary, vs. Site19/Coast/Bench)\nSurveyor P   0.6 Tide                58% / 31% / 11%\nSurveyor Q  −1.8 Bound                22% / 9%  / 69%   (toward Bench — Q's own naming-Coin\n                                                          is Bench-Sill-cast)\n```\n\nSurveyor Q's casting-spring is Bench Sill. The jar, witnessed by Q, leaned hard toward Bench. Witnessed an hour earlier by P — whose own Coin carries no strong kinship to any single spring — the jar gave a comparatively even spread. Two more pairings since have held the same shape: **the jar's reading correlates with the Witness's own provenance, not with anything in the jar.**\n\nThis is not the ordinary Kinship Assay bias from Addendum C. An ordinary unwitnessed gauge defaults to *its own* birth-water in the absence of a fourth anchor. This jar is *witnessed*, properly, every time — and still returns the Witness's signature instead of a stable account of itself. The working account: the jar was never cast from any spring at all. It has no birth-water to default to, kinship-wise, which the Assay already confirmed (it matches none of the three references). Lacking a self to compare against, and given a Witness who *does* have one, it does not anchor the reading. It **borrows** the Witness's.\n\n```svg\n<svg viewBox=\"0 0 360 160\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect width=\"360\" height=\"160\" fill=\"#f1e9d8\"/>\n  <circle cx=\"80\" cy=\"80\" r=\"40\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n  <text x=\"80\" y=\"84\" text-anchor=\"middle\" font-size=\"11\" fill=\"#3a2e1f\">JAR</text>\n  <text x=\"80\" y=\"135\" text-anchor=\"middle\" font-size=\"9\" fill=\"#3a2e1f\">no birth-water; empty kinship slot</text>\n\n  <!-- arrow from witness into jar, reversed from normal anchor diagram -->\n  <line x1=\"160\" y1=\"80\" x2=\"120\" y2=\"80\" stroke=\"#7a1f1f\" stroke-width=\"2\" marker-end=\"url(#arr2)\"/>\n  <text x=\"200\" y=\"60\" font-size=\"10\" fill=\"#7a1f1f\">Witness's own kinship</text>\n  <text x=\"200\" y=\"74\" font-size=\"10\" fill=\"#7a1f1f\">flows IN, fills the slot,</text>\n  <text x=\"200\" y=\"88\" font-size=\"10\" fill=\"#7a1f1f\">is read back as the jar's.</text>\n\n  <g transform=\"translate(280,80)\" stroke=\"#7a1f1f\" stroke-width=\"2\" fill=\"none\">\n    <path d=\"M-14 0 C-14 -12, -8 -17, 0 -17 C8 -17, 14 -12, 14 0 C14 6, 8 11, 0 11 C-8 11, -14 6, -14 0 Z\"/>\n  </g>\n  <text x=\"280\" y=\"115\" text-anchor=\"middle\" font-size=\"9\" fill=\"#3a2e1f\">Witness</text>\n\n  <defs><marker id=\"arr2\" markerWidth=\"8\" markerHeight=\"8\" refX=\"6\" refY=\"4\" orient=\"auto\">\n    <path d=\"M0 0 L8 4 L0 8 Z\" fill=\"#7a1f1f\"/></marker></defs>\n</svg>\n```\n\nProposed glyph, for a category the table has never needed before — not a Band, not QUICK's unbanded becoming, but a *vacancy that takes the shape of whoever fills it*:\n\n```svg\n<svg viewBox=\"0 0 160 90\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\" font-size=\"10\" fill=\"#3a2e1f\">\n  <rect width=\"160\" height=\"90\" fill=\"#f1e9d8\"/>\n  <circle cx=\"35\" cy=\"35\" r=\"18\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"1.5\" stroke-dasharray=\"3 2\"/>\n  <line x1=\"22\" y1=\"22\" x2=\"48\" y2=\"48\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n  <line x1=\"48\" y1=\"22\" x2=\"22\" y2=\"48\" stroke=\"#7a1f1f\" stroke-width=\"1\"/>\n  <text x=\"35\" y=\"65\" text-anchor=\"middle\" font-weight=\"bold\">UNKEYED</text>\n  <text x=\"35\" y=\"78\" text-anchor=\"middle\" font-size=\"8\">reads the reader</text>\n</svg>\n```\n\nIf Unkeyed wax exists, two uncomfortable possibilities follow. First: an Unkeyed gauge, mistaken for a clean one, would never show a stable target reading at all — it would show a rotating cast of whoever last held it, filed as \"noisy data\" rather than recognized as a different kind of instrument entirely. Second, worse: nobody knows whether Ashery's jar is a freak of casting, or whether *most* wax has a faint trace of this property and Site 19's enormous share of the foundry's output has simply been swamping it with birth-water bias strong enough to hide it until now.\n\n---\n\n## III. The Quickening Gauge, Pointed at Something Already Filed as Finished\n\nThe Quickening Gauge was built for a newborn. Nobody had tried it on anything the Survey already considered *settled* — until someone thought to run it on the M+R Yoke bead from Addendum E, the one that has sat at a stable −2.3 Cauls, Bound, since the season it was cast, and which both M and R asked never to have re-read aloud.\n\nThe standard gauge confirmed the same number it always gives this bead. The Quickening Gauge, recording continuously instead of stopping at \"settled enough,\" was left running for forty minutes — far longer than any prior trial, because nobody expected to need that long for something already filed as finished.\n\n```svg\n<svg viewBox=\"0 0 460 230\" xmlns=\"http://www.w3.org/2000/svg\" font-family=\"Georgia, serif\">\n  <rect width=\"460\" height=\"230\" fill=\"#f1e9d8\"/>\n  <g transform=\"translate(60,50)\">\n    <ellipse cx=\"80\" cy=\"0\" rx=\"14\" ry=\"70\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <ellipse cx=\"280\" cy=\"0\" rx=\"14\" ry=\"70\" fill=\"none\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <line x1=\"80\" y1=\"-70\" x2=\"280\" y2=\"-70\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <line x1=\"80\" y1=\"70\" x2=\"280\" y2=\"70\" stroke=\"#3a2e1f\" stroke-width=\"2\"/>\n    <!-- a VERY narrow band, almost a line, with a barely perceptible widening/narrowing wobble -->\n    <path d=\"M80 -3 Q140 -6 180 -2 Q230 2 280 -3\" fill=\"none\" stroke=\"#7a1f1f\" stroke-width=\"1.2\"/>\n    <path d=\"M80 3 Q140 6 180 2 Q230 -2 280 3\" fill=\"none\" stroke=\"#7a1f1f\" stroke-width=\"1.2\"/>\n    <text x=\"180\" y=\"-95\" text-anchor=\"middle\" font-size=\"10\">40-minute recording — M+R bead, \"Bound, settled\" per standard gauge</text>\n  </g>\n  <text x=\"230\" y=\"190\" text-anchor=\"middle\" font-size=\"11\" font-weight=\"bold\">\n    Residual Quicken Index: 0.04°/min — nonzero\n  </text>\n  <text x=\"230\" y=\"208\" text-anchor=\"middle\" font-size=\"10\" font-style=\"italic\" fill=\"#7a1f1f\">\n    far too slow for the standard gauge's settling window to register as anything but \"arrived.\"\n  </text>\n</svg>\n```\n\nThe band is not closed. It is narrower, by orders of magnitude, than the newborn's eleven-minute fan — but it is still narrowing, almost imperceptibly, forty minutes in. The standard gauge calls this Bound and stops looking, because Bound is what \"narrow enough, long enough\" has always meant to an instrument built to give a surveyor a number before lunch.\n\nThe uncomfortable implication, stated as plainly as Addendum G stated the Ceiling's: **the standard gauge may never have measured anything as actually arrived. It measures things slow enough to fool its own patience.** Three veteran surveyors' Bound Coins, the Yoke beads, possibly Site 19 itself sitting at the bottom of its own nineteen-year trough being read as \"true Still\" — every one of them may simply be Quickening too slowly for the old housing's settling time to catch the motion. Nothing has been shown to *stop*. Only to slow past the instrument's patience.\n\n---\n\n## Open Threads (for Addendum I)\n\n- **Is anything actually Still?** If the Quickening Gauge finds residual motion in a forty-minute trial of something already filed as finished, what does an hour find? A day? Has the Survey ever, in ninety years, held a target long enough to rule out a Quicken too slow for the standard dial — including Site 19's own original calibration?\n- **The Unkeyed hypothesis, scaled up.** Is Ashery's jar a singular freak casting, or a trace property latent in all wax, masked everywhere else by Site 19's overwhelming share of the foundry record? If the latter, every \"clean\" reading this log has trusted may carry a small Unkeyed component nobody has had reason to isolate.\n- **The positive Ceiling, untested by design.** The Coast-Sill Bound-floor experiment has a mirror nobody will run: a deliberately non-Site-19 Named-Twice attempt, to see if the fusion-point also moves with provenance. The Office's verbal position on this is reported as \"absolutely not,\" which the log notes without endorsing.\n- **Q's discomfort.** Surveyor Q, on learning the jar had read back Bench Sill through her specifically, asked to see her own naming-Coin record again for the first time in four years. She has not yet said what she found, and nobody has asked twice."
  }
]
Viewing live transcript · Export provides standalone page