Update on the serial number. It is not 18655 as I thought because this is on three parts, the frame, slide and barrel. The serial number is 36704 which is on two parts, the slide (inside) and the frame. The barrel does not have the serial number. 18655 is some kind of secondary number. This likely moves production into the first quarter of 1945 rather than 1943. On the steel, it is not Graf Spee armor plate. They did an analysis on them. It is however a really odd steel not usually used for firearms because it's too difficult to work. So theories it was provided by the Britishicon or the US are also unlikely. They knew what it would be used for. Which takes it back to the Graf Spee. Plenty of available steel there, just not the armor plate.

"For this article a Ballester slide, serial number 19924, well within the British-contract range, was sacrificed for chemical and spectrographic analysis. The results were both disappointing and perplexing. The slide turned out to be high-manganese low-carbon steel having 1.07 percent manganese, 0.33 percent carbon and 0.19 percent silicon, roughly corresponding to SAE1033. It contained virtually no chrome, nickel or vanadium. So much for the Armor Plate Theory ... .

While the metallurgical analysis does not show that the steel came from the Graf Spee, it also does not show that it didn’t. A warship contains large quantities of many different types of steel besides armor plate. Steel with high manganese content is not an ideal choice for gun manufacture because it is tough to machine. This suggests that HAFDASA was forced to improvise with raw material salvaged from something, but unless and until more evidence comes to light, its source remains a mystery."