# C1447 — Test MT_cumulative at Shem and Flood

## Question

Do the Flood-only, Shem-only and both-node partitions agree under the same accepted source state?

## Sources

```json
[
  "Registered MT_cumulative source",
  "Named600 hinge source review"
]
```

## Opened utc

2026-09-29T01:15:12.573909+00:00

## Results

```json
{
  "Flood": {
    "source_legs": [
      3430,
      9170
    ],
    "images": [
      3430,
      7190
    ],
    "path2": [
      1406,
      4836,
      12026
    ],
    "path3": [
      6,
      4106,
      7536,
      14726
    ]
  },
  "Shem": {
    "source_legs": [
      4030,
      8570
    ],
    "images": [
      3040,
      7580
    ],
    "path2": [
      1406,
      4446,
      12026
    ],
    "path3": [
      6,
      4106,
      7146,
      14726
    ]
  },
  "Both": {
    "source_legs": [
      3430,
      600,
      8570
    ],
    "images": [
      3430,
      600,
      7580
    ],
    "path2": [
      1406,
      4836,
      5436,
      13016
    ],
    "path3": [
      6,
      4106,
      7536,
      8136,
      15716
    ]
  }
}
```

## Finding

MT Cumulative repeats the LXX Regular response: the two coarse splits agree, while the retained600 refinement adds990. Its13016 refined head is inherited, not a correction.

## Reassessment

Test whether native LXX Cumulative shares that balanced response.

## Checks

```json
{
  "expected_heads": true,
  "same_source_total": true
}
```

## Closed utc

2026-09-29T01:15:12.574817+00:00

## Record provenance

Exact record hash: `804ee206785dd43655832b8fa6d4f9936dd2e45b913976afb5b41d6735966c34`

Source journal: `Research_Cycles/C1433_C1482/journal.json`