Walter Breen
The parallel to this state of affairs in the Philadelphia Mint is, of course, found in the "late Soho" and clandestinely restruck pieces allegedly originating in Boulton & Watt's Soho mint, dating from 1780's, but actually made in the mid 19th century. As Peck has told their story in voluminous and occasionally eye-twinkling detail, I shall not attempt to repeat it here. Restrikes were an attempt to pander to the collector - and commercial - impulse to have something nobody else has got; when the demand died down, at whatever cost, the supply did not continue to be renewed.
Early in the 20th century, apparently owing to the influence of the Paris Mint (which incidentally was more notorious for restrikes than any other in western civilization), medals and proof coins began experimentally to be made by a new process yielding no longer the old and tired brilliant mirrorlike fields, but instead a uniform granular sheen. Russell Nering has traced the process back to about 1896, in which year - if memory serves - Belgium and possibly some other small European countries issued a few proofs by the new processes. Britain introduced it on the Coronation proof sets of 1902, and Canadian sets of 1908. The Philadelphia Mint tried it on the St. Gaudens gold designs 1907-15, the Pratt gold designs of 1908-15, Lincoln cents of 1909-16, Buffalo nickels of 1913-16, and the extremely rare proof silver coins of new design of 1916, pattern and regular (these were not released to collectors in proof state). Somewhat similar proofs were made of the new Peace dollars in 1921-22, and of various commemorative half dollars through about 1935. Again, most of these (except for the fifty sandblast proof Hawaiian commemorative halves of 1928) were not allowed to get into collectors' hands.
Regrettably, the story here is not yet complete. At least five different variations on the matte and sandblast proof technique were experimentally used on regular proofs between 1909 and 1915. At the moment it is hardly possible to give verbal descriptions, or to tell exactly how they were made (the relevant records have not been released to the National Archives). What I do know is that the finish in each instance had to be applied after striking, and that in some instances it involved pickling the coins in weak acid and in others it involved spraying them with a stream of fine sand in compressed air. Russell Nering has experimented with the latter process and found that the fine and coarse sandblast finish, as on 1912-15 gold, evidently differed mainly in the size of sand particles used and in the velocity of the spray or blast. I do not know whether collector complaints were relevant or not, but the technique was abandoned by the British mint earlier than by the Philadelphia Mint: the 1911 proof sets and those of 1927 and later years revert to the old brilliant style, easily distinguished from uncirculated coins. It is thought that increasingly curved fields (spherical or paraboloid surfaces), being difficult to polish uniformly on dies, may have inspired use of matte or sandblast technique for proofs. Even this cannot be verified at present.
Proof that the matte finish had to be applied after striking is of some interest. I once owned a 1909 Lincoln cent with one side perfect matte proof and the other perfect uncirculated with the typical mint bloom found on all unc. and no proof coins. Under microscopic examination, it became obvious that the piece had been struck medallically (like all proofs of the period) but thatonly on the proofed side had the surface been interfered with. The type of surface involved made it plain that artificially roughened dies were not used, leaving only an after-striking treatment as tenable explanation. (Pickling or sandblasting before striking would have been obscured and replaced by typical mint bloom on impact, or by shiny surface were the dies polished.)
In any event, matte proofs are often simulated by uncirculated coins, and sometimes correct diagnosis is exceedingly difficult. Experience with the real thing is the only safe guide -preferably from proof sets of the year so that no mistake can be made - and even then this type of experience is of value only if the numismatist has an excellent visual memory. (Even then he is likely to develop ulcers making honest decisions on 1916 and 1909 VDB proof cents.) No verbal criteria exist for unequivocally identifying matte proofs, nor will photographs help; the language lacks words for the subtle distinctions involved.
Matte and sandblast proofs ceased being made for reasons never published, aside from World War I, but one reason probably was that these coins do tarnish much faster than regular uncirculated coins. The process used to make such proofs activates the metal surface, causing exceptionally fast oxidation.
When the Philadelphia Mint in 1936 decided to resume sale of proofs to collectors, Coiner's Department personnel had evidently forgotten how to make 19th century style brilliant proofs with mirrorlike fields and frosty devices. Proofs of 1936 thus come in two separate styles or issues, both with typical polished edges, sharp rims and borders and well brought up devices. The first issue has surfaces very much duller than uncirculated coins, without mint bloom or the granularity of sandblast proofs; they are nearest the "satin finish" style of some proofs of 1907, 1909 and 1916. The nickel and cent of this issue come usually much darker than uncirculated coins. On the other hand, the scarcer second issue proofs of 1936 have almost the entire die surfaces mirrorlike, as in 1937. These are among the most brilliant proofs of the period. Later proofs through 1942 are much closer to the 19th century style, save that frosty devices are very unusual. Nickel proofs have little advantage in sharpness over the regular uncirculated coins. Much the same comment can be made for proofs from 1950 to the present; the first proofs of 1950 were quite dull, frosty devices are seldom encountered, and nickel proofs leave still a great deal to be desired in sharpness, as do sandwich-metal proofs from San Francisco, 1968 to date.
Modern proofs (or specimen coins) issued by other mints - Canadian, British, and Continental - for the most part also have resumed the typical 19th century appearance, and not much need be said about them. Proofs represent the Mint Bureau's sole lagniappe for present-day coin collectors; they represent a source of needed revenue for "postage-stamp" countries and others; they represent, even now, supposedly the finest work modern minting methods can produce, within the limitations of modern coin design, which are the limitations of low relief (to facilitate stacking) and stereotypy (to frustrate counterfeiters), which are ultimate outgrowths of the experiments of the key figure in modern minting technology - Matthew Boulton.