# Nearby radio activity does not identify the best infrared counterpart

13 October 2026. The frozen three-case comparison is independently verified:
**0/3 fixed best infrared records** have a qualifying original MMB maser position
within one arcsecond. **None of the six retained infrared alternatives** has any
native catalogue maser position within that radius. The predefined projected-
geometry criterion is not met. This is no physical starlessness result or
exclusion of embedded protostars.

The first field has useful, separate positive context: historical radio maser
**330.878-0.367** is **12.520 arcsec**
from the dust-clump centre, but **17.501
arcsec** from the best infrared point. Regional molecular/radio activity cannot
be assigned to that point merely because the region received a high clump rank.

## Same cases, native coordinate fields

| Clump | Fixed best infrared record | Nearest native maser name | Point-to-maser, arcsec | Clump-centre-to-nearest-maser, arcsec |
| --- | --- | --- | ---: | ---: |
| HIGALBM330.8814-0.3656 | SSTGLMC G330.8829-00.3661 | 330.878-0.367 | 17.501 | 12.520 |
| HIGALBM351.2258+0.6937 | SSTGLMC G351.2258+00.6940 | 351.242+0.670 | 104.881 | 103.458 |
| HIGALBM353.4272-0.3734 | SSTGLMC G353.4273-00.3723 | 353.410-0.360 | 76.937 | 79.066 |

For the third field, the distinct red alternative's nearest radio record is
81.837 arcsec away. It is still separate from
the blue best counterpart and its time-series measurements; neither gains a
close maser association. Every alternative, nearest radio record, native flag
and pair within 1/15/30 arcsec is retained in the source assessment.

The earlier cone is bound by its [original query registration](../artifacts/native_glimpse_registration_20261012.json);
this metadata is separately identified as post-read interpretive context.

The first radio record is native ordinal string `VIII/96/catalog.dat:000307` in the
hash-bound historical catalogue, at literal J2000 coordinates
**RA 242.582500000 deg, Dec -52.102194444 deg**. Its original
reference field is retained. Radio positions describe emitting gas; ordinal/name
aliases are not identifiers of independently confirmed stars. Epoch-dependent
activity, physical membership and stellar identity are not measured here.

## Post-read coverage qualification and interpretation

The 12.520-arcsec first-field offset is just outside the earlier **12-arcsec
GLIMPSE query around the clump centre**. This is a post-read coverage observation,
not a new primary criterion, changed search radius or substitute source. The old
catalogue selection/negative gates remain frozen. A catalogue counterpart outside
that original cone was not tested in the earlier infrared selection, and this
unit has not acquired one. The first field has a nearby published radio marker;
we have not established that it belongs to this Hi-GAL clump or caused its
molecular activity or the matched infrared point's temporal measurements.

Neither the 15/30-arcsec contextual circles nor a large molecular beam is a
physical clump boundary. This potential confusion explanation still needs a
separate source-specific test. Positional proximity alone cannot attribute
emission, establish a star's evolutionary stage or count a discovery.

Radio absence at an infrared point does not exclude young stars: the historical
maser sample traces a selected activity window, has finite sensitivity, and gas
can be displaced from the star. This 707-record I–IV snapshot is not all modern
masers. Blank position flags supplied only the registered geometric eligibility;
no quoted position-error or source-epoch contract was available. We performed no
frame/proper-motion transformation, calibrated probability or parental-velocity
membership comparison. Native velocity/flux/reference fields remain exact context.

## Independent verification and resources

The protocol, code, exact cases and synthetic controls were frozen at
`bc23a8c7249088f01583ebbc0f678b627f31be83` before any native catalogue position readout. One public static-file
GET returned **69,855 bytes / 707 records**, with no coordinate query/upload,
retry, redirect, image/cube request, paid compute or new survey. Every registered
existing input remains hash-identical. The terminal NEOWISE request, N20 and
c6828 remain closed; none was restarted or repaired by this radio comparison.

The worker used Astropy angles and separations. The independent checker imports
no worker/Astropy code: direct raw slices, Decimal sexagesimal conversion and
vector cross-product distances recomputed **all 707 records and 6,363 angular
pairs**, native flags/fields, nearest/radius lists and the primary decision.
Maximum coordinate difference is 1.14e-13 degree,
a software consistency value, not scientific positional accuracy.

All four guarded stages (synthetic registration controls, acquisition, readout
and independent check) exited zero without a resource stop, each in about one
second, with at least 62.40 GiB observed free disk.
Their short lifetimes mean RSS samples miss possible transient peaks; do not
claim small actual peak memory from them. Each stage is below 1 MiB and the
complete code/data/report/receipt packet below 4 MiB. Original inputs are preserved.

Evidence: [frozen protocol](NATIVE_MMB_POSITION_PROTOCOL_20261013.md),
[registration](../artifacts/native_mmb_position_registration_20261013.json), [complete original catalogue](../artifacts/native_mmb_position_acquisition_20261013/catalog.dat),
[all parsed native fields](../artifacts/native_mmb_position_analysis_20261013/radio_records.json.gz),
[source assessment](../artifacts/native_mmb_position_analysis_20261013/summary.json),
[independent check](../artifacts/native_mmb_position_verification_20261013/verification.json),
[final checkpoint](../artifacts/native_mmb_position_checkpoint_20261013.json). Primary schema/reference sources are linked
in the frozen protocol; original names/flags and repeated native instances remain intact.

## Position against the full goal and stop

The measured conclusion is source-specific: nearby historical radio-marker
context in the first field does **not** coincide with its best infrared record
at the frozen scale. The other best/alternative points also lack such a match.
This qualifies the use of coarse molecular activity and clump scores as apparent
support for a particular point source. It supplies no independent held-out youth
label, physical class or physical negative. MMB references and case selection are
consumed, and the global 19,328-parent/19,066-unknown ledger remains unchanged.

All original models and gates stay fixed: embedding 14/27 versus nonlinear 15/27
at 18 reviews on consumed chemistry proxies; failed original Spitzer 0/3,
cadence-majority 3/27 and uncertainty 333/800 gates. The new radio criterion is
distinct from them. Infrared occurrence remains unmeasured, hypothetical 814
independent-bin support unvalidated, and 9–25%/45–55% scenarios descriptive
sensitivity rather than confidence bounds or lifetimes. No new star, stellar
birth, physical precision/recall or robust detector is established.

This unit stops with its verified projected comparison and qualifications.
No expanded infrared cone, source substitution, relaxed match/flag criterion,
new fit or automatic successor experiment follows. The full goal stays active
and unmet. Results are local until a separate publication/deployment receipt
exists; no private Atlas action is authorized.