The photos of the x-ray equipment are interesting but based on information derived from the Pop. Mechanics article and a little number crunching, the caption information seems a little dubious to me. But, admittedly, I know nothing about metallurgy.
First, 1/1000000 of an inch is equal to about 25 nm which is roughly the size of molecules. An x-ray's wavelength is about 10 nm and perhaps that is where the translation from scientist to reporter went awry. For the article to imply a crack of 1/1000000 of an inch could be visualized is pretty ludicrous, at least in 1937. I might be mistaken but I don't think molecules and atoms were visualized until the current century.
Further complicating this supposed resolution is that even modern film with the best apochromatic lenses can only resolve around 150 lp/mm (line pairs per mm) )under ideal conditions. (Just a few of the limiting conditions would include, shutter speed, stop, temperature, humidity, processing chemistry, and even the inherent quality of the imaging equipment.) At 150 lp/mm, one lp is about 1012 nm wide, significantly larger than 25 nm.
I don't doubt that the x-ray equipment of 1937 could resolve small cracks but not anywhere near the size spoken of in the Pop. Mechanic article.
cheers,
david alan