UAP AnalysisIndependent · the declassified record
← All incidents
Dept. of DefenseUnknownAmbiguous

Project SIGN soil sample analysis — Wright-Patterson / FBI Lab, September-October 1948

Sep 9, 1948Location of original depression not specified (FBI Lab analysis, Washington D.C.)
⟳ Deeper read pending. This record is from the FBI 62-HQ-83894 flying-disc file, whose cursive, clippings and faint scans are under-read by automated OCR. It is flagged for a vision-transcription pass and likely undercounts what the source contains — see the documents collection note.
Analysis — our summary

Section 4 contains the FBI Laboratory's formal analysis report on the soil sample submitted by Project SIGN (Wright-Patterson AFB) on September 9, 1948. The sample was from a depression allegedly made by a landed flying saucer. FBI Lab examination was microscopic (petrographic-geologic). The formal report was completed by October 7, 1948. No surrounding soil comparison sample was submitted, limiting comparative analysis. The ultimate laboratory findings are not fully recoverable from the available OCR text.

As reported — verbatim from the document
Project 'SIGN': soil sample from depression approximately 2 ft diameter, 1 ft thick, made by object that 'settled gently, rebounded to 20 ft'; FBI Lab: 'Examination by JEVONS... [result text garbled by OCR]'.
Analyst notes — caveats & confidence

Section 4 file was read to 1,499 of 4,742 total lines; lab result details are partially OCR-garbled. The same incident is referenced in Section 2. Lab examiner initialed 'JEVONS' is recoverable; final determination not clearly readable. No geographic origin of the depression is specified in either section. Confidence lowered due to truncated read and OCR quality.

Provenance
Source document65_HS1-834228961_62-HQ-83894_Section_4.pdf
Document typecase file compilation
Reporting agencyDept. of Defense
Source pages214
DeclassifiedFirst public at this release (2026)
Held classified~78 years (≥, to this release)
Extraction confidence LowHow cleanly this record could be parsed from the source — driven by legibility & redaction. It is not a measure of how credible or anomalous the sighting is.