# EX-15: replicated transmitted dependency

EX-15 is a replication rung, not a search for a new mechanism. Keep the exact
EX-11–14 ecology, resource barrier, mutation operator, rates, population size,
horizon and controls. Use fresh discovery seeds only; do not seed B, BC or any
successful genotype. No parameter is changed to increase opportunity supply.

## Replicated unit and attrition

The independent unit is a separately discovered full-world B→C opportunity, not
an assay seed. Log denominators at every transition:

`B discovered → clean B→C candidate → birth-time opportunity supported → BC
transmitted → intact BC descendant at checkpoint → later dependency supported`.

The first clean B→C candidate in each full world is selected online before
outcomes, with no survivor substitution. Controls remain full, B-only,
no-substrate and atomic-only, ten paired seeds per arm for each discovery batch.
Continue discovery batches until either three independent cases pass the complete
pipeline or a fixed maximum of 100 full discovery worlds is reached. The maximum
is a stopping rule, not a license to inspect and choose favorable worlds.

## Per-case evidence

For every candidate, preserve the actual pre-B background and serialized
birth-time ecology. Birth-time opportunity requires B>A, BC>B, C-only<=A and
positive interaction, each in at least8/10 paired futures and in the registered
mean-sign direction. Use EX-12's two contexts separately; do not pool contexts.

Transmission requires actual inheritance of exact BC by a descendant. Persistence
requires an uninterrupted BC descendant at the fixed +200 checkpoint. Historical
event lineage and functionally equivalent BC elsewhere remain separate identities.

For each supported transmitted case, repeat EX-14's 100-future checkpoint assay.
Report unconditional `N_g` and `I_g=1[N_g>0]` for A, B, C-only and BC over all
100 paired futures. Separately compute `D=N_BC−N_B` only where `I_BC∨I_B` is
true. No zero/zero future is removed from unconditional realization. Conditional
effect support requires D>0 in at least80% of realization-event futures and a
positive mean; fewer than8 events is unestimated.

## Primary success

EX-15 supports replicated retained depth-1 dependency only when at least three
independent opportunities pass every transition through later conditional-effect
support. Report the complete attrition table even if fewer than three pass. Zero
passing cases is a completed negative/inconclusive outcome, not a reason to alter
the maximum, rates or gates. Accumulation beyond B→C, historical necessity of B,
environmental mediation and D remain untested.