Documents / Report
The Roswell Report: Fact versus Fiction in the New Mexico Desert was published by Headquarters United States Air Force in 1995. The Black Vault distributes this copy. It reproduces the report by Col. Richard L. Weaver and the synopsis by 1st Lt. James McAndrew, both written after a General Accounting Office inquiry requested by Representative Steven Schiff. The Air Force search found no evidence of an extraterrestrial craft or crew. It concluded that the Roswell debris most likely came from NYU Flight No. 4, a Project MOGUL balloon train.
“Gross”27 pages
properly at 12,000 feet, but the air entrapped in its aneroid was heated and caused the operating level to be at the lower value. This would correspond to a superheat of $ 3 0^{\circ} \mathrm{C} $ above the air temperature.
Later flights showed that the type of load attachment used on this balloon was unsatisfactory; however, with proper rigging, cells of .008" thickness were good vehicles as they usually showed very low diffusion and gas leakage.
Near the end of the recorded data, the height-time curve shows large oscillations about a pressure plane (Figure 6). Three factors which probably
Figure 6
contributed to this instability were:(1) the turbulent motion of the heated air over the desert, (2) the changes in temperature of air in the aneroid valve as intermittent clouds shut off the sun, and (3) the overcompensation caused by the valve-controlled ballast flow.
On this flight the first "destruction device" was used for the purpose of bringing down the balloon after a fixed time to prevent excessive interference in air-traffic lanes. This particular model was a clock-driven device which failed to operate, probably because of low temperatures causing unequal contraction within the movement. Its action was to consist of detonating an inflammable compound taped to the balloon, rupturing its side and permitting a rapid escape of the lifting gas. Report, cited by the archive. The PDF is mirrored here; the original link is under it. The text was read from the page images by an OCR model; expect the odd misread word. 882 pages are in the text index: search them above, or from the library's search.