Although the first Spanish proof house was established at Eibar in 1844, until 1923 proof was voluntary. After 1923 it became compulsory. This had something to do with those annoying international agreements which plagued Europe after World War 1, and just gave everyone headaches without actually doing anything useful to avoid the big problems .... big problems like, say, the approach of World War 2.
The Spanish proof stamps changed in the very late 1920s and very early 1930s, so there were what I'll call "early" and "late" versions of interest to us here.
And finally, the letter codes for proof years -
So, just when did the Eibar proof house switch from the older stamps to the newer ones? It looks like the big year was 1928. Guns proofed in 1928 might have either the older or newer stamps. Of course dating the older stamps can be an exercise, as they don't have a date code. Other information such as the serial number of the gun must be used to estimate the manufacture date, which puts limits on the proofing date. Obviously, errors are possible. But so far, 1928 is my best guess for the transition date.
Everything above is general for Spanish handguns, and applies to Beistegui Hermanos guns as well as Unceta.
Now for photos of actual specimens. I have blown up most of the photos so that they're of uniform scale. The penalty for that is that some of them get fuzzy. But since all we're interested in is proof marks, the fuzziness of other features isn't a huge drawback.
Proofs on the Astra 900 Series
Astras were generally proofed. Some strange things happened during the Spanish Civil War, when significant numbers of guns were either stolen or "stolen" while in storage, and just who did the stealing rather depended on exactly who - the Republicans, the Nationalists, or the (for a while) independent Basque government - controlled Guernica at the time. Those guns may or may not be proofed. And there was more than a bit of outright "requisitioning" going on, starting with the Republicans about 1931, and continuing until after the end of the war in 1939. And sometimes guns shipped directly to a Spanish military or police organization weren't proofed, perhaps on the theory that the military would do its own testing of the product before acceptance, something not practical for civilian customers.
So, with the understanding that sometimes there were no proofs at all, the usual places for proof stamps on 900 Series guns were -
A pair of single proofs on the right side of the barrel and barrel extension, on either side of the joint between the two. In contrast with standard Mauser practice, these were two separate parts on 900 Series Astras.
A triad on the right side of the frame. The exact location of the triad varied by model and production date.
The Eibar proof house showed a preference for stamps on the left sides of guns, all things being equal. But on the 900 Series Astras things weren't equal - most of the left side of the frame wasn't frame, but a sideplate. So on the 900 Series the proof house put the stamps on the right side. It appears that after World War 2 the proof house went to more effort to standardize, and stamp all guns on the left side - and sure enough, places for the stamps were found on the left sides of the last 900 Series guns. But earlier stamps were all on the right side.
Here is a prototype Model 900, serial number 4, after Antaris. The triad is just forward of the trigger pin. It's a bit hard to make out the first stamp of the triad, but it's STAMP 3, the crowned shield with saltire couped, followed by P.V., with the lion rampant bringing up the rear. In other words, the standard "early" triad. The lion rampant is also on the barrel, just forward of the chamber area. On the barrel extension, the lion rampant is much further aft than seen on production specimens - it's just forward of the sight pivot. This gun is stamped 1927 on the frame, hidden under the left grip panel.
This next is an early production Model 900, serial number 248. All proof stamps have moved to their normal places for production 900s. The triad has moved aft a bit, to the space between the barrel locking pin (the square one) and the trigger pin. It's obviously the "early" triad, and the lion rampant is on both the barrel and the barrel extension. This gun has the small Bolo-style grip, solid safety lever knob, and small ring hammer - all early features.
Serial 1768 has a later feature than the above guns - the large grip. It still has other early features, though - the fine grip panel grooves and small ring hammer. The triad hasn't moved and is still the early type. It's the one I show in closeup above, as an illustration of an early triad. The frame of 1768 is stamped 1928 under the left grip panel.
The pair of single proofs on the barrel and extension of this gun show well in closeup.
Sometimes we have to do some elementary detective work to tell what's going on. Here's serial number 2672, a very similar gun to 1768 - large grips, fine panel grooves, small ring hammer - with the same triad, but it's hell to read -
We can't say for sure that that's the early triad on there unless we can figure out if the illegible figure is the familiar lion rampant. Fortunately, the single proofs show clearly on the barrel and extension of this gun, and since we know they're the same as the third stamp of the triad, we can be certain that it is indeed the standard early triad.
Here's a late triad and, naturally, late single proofs on the barrel and extension, all on a Model 901, serial number 15596.
And now, a weirdo gun with normal proofs. Erickson & Pate, page 231, shows this 20 shot fixed-magazine semi-auto, describing it as a factory special Model 902, serial 26705. Unceta seems to have made two small batches of semi-auto 902s, which makes them 900s with longer barrels and larger magazines. The factory box says Model 900. But no matter - the proofs are ordinary late single proofs on the barrel and extension and an ordinary late triad, all in the usual places. The year code in the center of the triad is an H (wish a better photo was available - the star over the H may actually have been struck decently this time), so this was proofed in 1935.
This Model 903, serial 26760, has the selector switch (as it should). Somebody decided to avoid the switch area altogether by stamping the triad down just over the grip. The late single proofs on the barrel and extension were left in the normal places.
Serial number 27940, a slightly later Model 900 (and of course the Model 900 never had a selector switch), kept the triad in the new location, down above the grip.
My word, things can get crowded down there. Here's a fairly late Model 900, serial 29096, with a late triad (as expected) over the grip (as expected), but there are some other stamps too - some sort of crossed scepters with three letters at left, and a circled BV at the top edge - Birmingham View proof?
Here's the same gun, with more stamps on the left of the barrel extension, near the ejection port -
It's recognizable now - another triad, this one from the Birmingham proof house. So at some time this gun was imported to England and reproofed, as England in those days (pre-1980) didn't trust "funny foreigners" to proof things the proper way. This triad is BV (the Birmingham view proof - basically, a final visual inspection after the proof firings), BP (originally, the Birmingham black powder proof), and BN (appearing in conjunction with BP, this is the Birmingham "semi-smokeless" proof - what "semi-smokeless" is, I've yet to figure out). The crowns above the letters have to do with England still being a monarchy, and all. The circles were code for foreign-made guns. To be sure no one missed that, they'd also stamp NOT ENGLISH MAKE on the gun. I believe that's what the text to the right of the British triad says in the photo above, hard though it is to make out. Here's what the Birmingham triad looks like on a gun of impeccably English (not to mention Birmingham) pedigree, a Webley Mk. 4 revolver made in the early 1950s -
So the circled BV on the right side of the Astra is just another view proof. The crossed scepter thing down just above the grip still has me baffled, though. Birmingham used a similar mark for a while, but it was long obsolete by the time this gun was made.
Which brings up another question - just when was this gun made? The F year code tells us that she was proofed in 1933. However Unceta production records were fairly good for that period. And they show that serial 29096 was part of a batch of 626 Model 900s completed in 1934 (numbers 29001 through 29628, skipping numbers 29626 and 29627, which were completed as Model 903s). The question is, how did Unceta define completed? If guns weren't counted as completed until after they were fitted and assembled, shipped to the Eibar proof house, proofed, shipped back, and whatever final finishing steps were performed (I don't know, maybe they added the lanyard rings last), then it is possible for a gun to be proofed in 1933 but completed in 1934.
A further example, this time one of the very last examples of the Model F, serial number 30796, made (like nearly all Model Fs) in 1935. Nombre de Dios! The triad has moved again, though not very far this time. The Model F had the cyclic rate reduction mechanism inside the grip, and apparently Unceta integrated that mechanism with the selector switch. Whatever the excuse, the switch moved aft, from directly behind the trigger to directly over the grip. So, sensibly, the proof triad swapped places with the switch. Now, if they'd left the triad up between the barrel and trigger pins, near the top of the frame, they wouldn't have had these problems. The triad is, as we'd expect, the late type, as are the stamps on the barrel and extension. I'd make a tentative guess that the year code is H, which would mean she was proofed in 1935 - the same year she was manufactured.
Peculiarities show up. Number 31808 was a Model 903, with a year code K, indicating that she was proofed in 1938. Unceta production records are a bit vague at this point, but they appear to show that the factory recorded her as completed in 1939. She was one of the thousand 903s sold to Germany in 1943. All very typical adventures for a late production gun. But her proof triad was stamped in the old position, up near the top of the frame, rather than above the grip. So although the proof triad tended to migrate downward and aft with time, it sometimes reverted. But the excitement wasn't over yet - the triad made one last move.
While we've been admiring the right side of the 900 series frame, there's been deviltry afoot. Consider the circled areas of these specimens:
Hmmm. On the earliest guns, the frame area aft of the triggerguard had a generous blend or radius. It doesn't show in the above photos, of course, but the radii were symmetric - identical on both sides. As production progressed, there was a tendency to decrease the radius, until it was entirely gone by the 30 thousand serial range - the photo of the Model F. That was actually useful on the Model F, as a shaft - part of the cyclic rate reduction mechanism - extended through the frame, with the ends visible from both sides. But the serial number couldn't be stamped on the end of the shaft, so it had to move forward a bit. With no frame radius there, that was no problem - plenty of room for a serial number to crowd forward.
The radius seems to have crept back a bit on the latest Model 900s, some three thousand serial numbers later, and the serial number moved back to where that pin wasn't (since the Model 900 had no use for the cyclic rate reducer). But then it disappeared again on the Model E. This is not really a surprise, as the Model E frame was directly derived from the Model F, as was the magazine. Without a Model 904 or Model F-type cyclic rate reduction mechanism, that shaft end sticking over to the left side went away, and the serial number could move back to its usual place directly over the grip. But that left a flat piece of frame directly in front of the serial number, and the proof triad ended up there. The single proofs on the barrel and extension also moved to the left sides of each.
So on the Astra Model E, all proofs eventually ended up on the left side of the gun. I believe that sometime during that long gap in 900 Series production - from 1941 to 1949 - the Eibar proof house decided to standardize the locations of proof stamps, and since most guns proofed there were stamped on their left sides, that's where they'd stamp those last 900s as well. Here's a view including all the stamps on a Model E.
It looks like the order of the stamps in the triad changed, also. I see at left the Admission to Proof mark (the saltire couped with shield and helmet), then the final proof (the P-bomb), and finally at right the year code. That year code could be an X, which would be 1952. According to the Astra production figures, serial 34319 was probably made in 1951. In the world of Astra, that counts as pretty solid confirmation.
The mysterious Hope stamp, which appeared on the barrel or top chamber flat of the earliest Model 500s - up to serial number 500 or so - was accompanied by what certainly looks like a proof mark - a B G surmounted by a crown.
Here are photos of three different guns (one photo each from Erickson & Pate, Antaris, and Gun's World), suitably scaled by me.
Yes, they're different guns, all right, not just different photos of the same gun. But to me all three Hopes, including the wannabee proof, look identical. That means that the text of Hope and the wannabee proof were on the same stamp, or die, or silkscreen, or whatever Unceta used to put it there. So unless the word "Hope" was part of the proof mark (not too likely - the whole idea of a "proof" is, by actual test, to prove that the gun won't blow up, not to hope that it doesn't), the mysterious crowned B G isn't a proof mark.
Proofs on Beistegui Hermanos guns
Documentation (meaning, mainly, photos) is slim on this topic. The guns were indeed proofed, and with the same single proofs and triads we've just seen all over the Unceta product. However, locations were a different. Based on a limited number of samples, I'd say that the First Model guns were stamped on their left sides. Second and Third Model guns were stamped on their right sides, with the triads just above the grips.
This first example is from E&P, page 214, where it is captioned "Royal". But it is clearly marked ETAI on the right side of the frame, and not marked Royal at all. This is one of the First Model Beistegui pistols, an early (semi-auto only) variant. But it can't be too early, as we can see the later style single proofs (the "P" bomb) on the barrel and barrel extension, and the later triad on the frame. Although it would take a genuine man of faith to be sure of the proof year in the fuzzy photo, it could be a B, for a proof date of 1928. That would be consistent with the serial number of this gun, 12713 (maybe), which would make us suspect that she was made in 1928. The proof stamps are all on the left side of this gun.
Here's a nice murky photo of a ROYAL-marked gun, an early select-fire version. The proof triad is barely visible in the same location as on the semi-auto ETAI above. The single proofs on the barrel and extension are invisible here, as is the proof year code.
Here are a couple of select-fire Second Models, both marked MM31 on their left sides, with no brand marks on the right. No magazine release buttons are visible, so they must have fixed magazines, and so must be First or Second Variants. All proof stamps have moved over to the right sides. Perhaps by this time the Eibar house, stamping Astras on their right sides to avoid the sideplate on the left, had decided to stamp all Mauser-esque guns, regardless of manufacturer, on their right sides. There was plenty of room on the left side of the MM31, even with the selector switch in the way, if anyone had really wanted to put the stamps there. In any event, the proof stamps are the newer type, and the year code, D, means that both guns were proofed in 1931.
The next exhibit is a Fourth Variant of the Second Model, a select-fire gun, serial 31632, with a magazine interchangeable with Mauser Schnellfeuer magazines, and supposedly sold as a Super Azul. I don't see any brand markings on it at all. The year code, H, means she was proofed in 1935.
This is one of the third and last model of pseudo-Mauser made by Beistegui, an MM34. It is marked as such on the left side of the frame. The late-style proofs are still on the right side. The year code is J, so she was proofed in 1937.
The proof stamps never moved back to the left sides of the barrel, barrel, and frame on the Beistegui guns, as they didn't survive long enough. The Astra Model E had proof stamps on the left side, but the Model E didn't come along until a decade after Beistegui Hermanos stopped making guns.
Leonardo Antaris, Astra Automatic Pistols, has some good, though incomplete, material on Spanish proofs.
Gerhard Wirnsberger, The Standard Directory of Proof Marks (translated by R.A. Steindler) has a lot of material, if you have the fortitude and energy to dig it out.
When it comes to proof marks, something not shown in the books always shows up. And what is in the books, often doesn't quite match what's actually on the guns, particularly dates of use.