Streak candidates on pre-Sputnik Hamburg plates and a test of orbital recurrence

Narrow streak-like marks on Hamburg photographic plates exposed in August–September 1953 provide a case study in archival recurrence and tests of common motion. Beginning with features on consecutive August nights, we combine native-image measurements, rotated scans, orbital reconstructions and controlled searches of other exposures. Both August marks persist in repeated scans. Among 536 same-field plate regions, the original 33-day episode is the only qualifying five-night set within a 60-day window; a second enters at 62 days. Selected August and September subsets admit continuous trajectories, and a two-impulse construction connects five selected points. Synthetic histories show that similarly close fits can also arise from unrelated marks. Extending the September 15 measurement to the full visible trace reveals curvature that all six saved constructions fail to reproduce, with discrepancies reaching 79 arcsec. Further August checks use 148 Gaia-associated stars: the marks remain stable under stellar registration, and subtraction of a neighbouring stellar image leaves structure along the August 14 trace. The tested profile statistic has insufficient discrimination to establish optical formation. A targeted Vatican search over 36 million additional native pixels yields one short on-path trigger, reproduced as stellar structure on the preceding exposure. The August pair remains unresolved, with reproducible image measurements and explicit predictions for further examination. Establishing whether its marks formed during the recorded exposures, and whether independent observations support common motion, would determine their astronomical significance. Version 2.7 expands the original August-pair study with regional recurrence searches, dynamical and synthetic-history tests, literature comparisons, and image-formation controls. This research preprint has not been peer reviewed. The deposit includes the paper, its complete LaTeX source and analysis bundle, and the original research supplement. The manuscript's package map also documents 21 separately prepared archives that are not included in this deposit. The source bundle preserves the reviewed manuscript and numerical provenance. Historical exploratory reports inside the archives are retained for reproducibility; the v2.7 paper gives the current interpretation. The paper and original analysis are licensed under CC BY 4.0. Third-party archival material retains its source terms and credits; see the manuscript and archive documentation.