It is generally believed that the sale of proof sets to individual collectors, at a premium over face, began in 1858 under Mint Director J. R. Snowden, though the card accompanying cased sets in the 1840's may tend to indicate that some limited sales did occur in that decade. This view is confirmed by a note (courtesy of Raymond H. Williamson) quoting George F. Jones's Coin Collectors' Manual of 1860: "The 'Proof Sets,' so called, contained the Silver Dollar, Half Dollar, Quarter Dollar, Dime, Half Dime, Three Cent Piece and Cent, and have been issued regularly since 1840, with the exception of the year 1853, in which no proofs were coined. Prior to 1860 (i.e. 1858? WE) they were given to collectors, from the Mint, for $2.02, now they are $3.00." There are several remarkable things about this contemporaneous notice. Not only does it support the view that cased proof sets were sold in limited quantities in the forties, but it also, accounts for the extreme rarity of proofs dated 1853. The "commonest" 1853 proof coin today is the silver dollar, of which twelve were struck (all being restrikes); only five sets were apparently struck of silver coins with arrows, and no copper or gold at all!
There is no other record of a sale price quoted for silver-minor proof sets prior to the Snowden broadside, of which more below. Face value 1840-50 would have been $1.91 1/2 1851-January 1857, $1.941 1/2; May 1857 through 1860, $1.94, so that the $2.02 evidently included 8¢ proofing charge. The commonness of the large cent proofs of 1857 in comparison with the Flying Eagle cent proofs suggest that the Mint's regular customers had bought their yearly sets as soon as word got around that they were ready for purchase. As in the 1840's a charge of over $3 was made for a display case for gold proof sets, presumably the cost of the case for a silver-minor set would have been nearly as high -likely doubling the cost of the set! We need look no further, then, for a reason for the absence of original display cases for silver-minor proof sets prior to the eighteen sixties. Few collectors would have been willing to pay double the normal charge merely to have their coins in cases, especially when (as often happened) they broke up the sets and included the coins in year sets of the various denominations.
The above begs the question of when the sets of coins acquired the "proof". So far as I know J. R. Snowden appears to be the creator of the term, or rather the one who first applied it to what had formerly been called "master coins." R. W. Julian's discovery of a record of five proof gold dollars of large size or second design being struck in 1854 brings the use of that term to within the first year and a half after Snowden's accession to the throne in the Philadelphia Flying Disc Manufactory, but its first public use known to me is in Snowden's broadsides of 1858 and later years giving "regulations" to the public respecting transactions with the mint. Leaving aside the provisions therein respecting redemption of obsolete coins (e.g. Spanish silver, old copper cents and half cents, etc.), the relevant material is the sections establishing the price to the public of proof sets. "Silver sets" (cent to dollar, face $1.94) sold at $3.00 apiece, "gold sets" (gold dollar to double eagle, face $41.50) at $43.00, complete sets $46.00, without mention of display cases. As original sets from the eighteen sixties survive with fair frequency in identical disintegrating paper wrappers slightly thicker than tissue paper (cheap sulfite paper), the presumption is that the sets were sold directly to collectors in such wrappers, as they certainly were in the 1892-1915 period. Snowden may well have realized that many such sets were destined to be broken up and added to growing year sets by denomination, making permanent cases a needless frill. However, at least some sets from the eighteen sixties did leave the mint in specially made display cases. I have seen silver sets (cent to dollar) of 1865, 67, 68, 70 in such cases, generally of buckram or morocco in some indefinite color which may once have been maroon or plum but is now faded to mud, lined with dark blue plush (the inside upper half sometimes being satin), and containing spaces for one each of the coins of the year. In at least two instances a recipient's name was gold stamped on the outer cover. The 1867 set had room for only one nickel - that without rays - and it was presented to its recipient in March 1867, requiring revision in my earlier estimate of the number of proof nickels with rays. (See below, under 1867). The presumption is that cased sets were made only rarely on special order, if indeed the cases came from the mint (and all I have seen are much the same as later ones which definitely originated in the mint), and included only the proof coins currently being made. In other words, a proof set issued in 1864 would have included only one cent, whichever type was then being issued, and only one 2¢ piece - the latter only if issued after April 1864. Any set including both nickels of 1867, or both 2¢ pieces of 1864, or 2 or 3 different cents of 1864, must have been partly or wholly assembled by the collector. Jon Hanson tells me that he has seen one such cased set lacking one of the regular denominations of the year because none had as yet been struck; it came from this same period, but he does not now recall the date. (I suspect it to have been 1866 lacking the nickel 5¢ or 1865 without nickel 3¢.) Any "complete" set of 1861 issued between April 5 and April 14 would have lacked the dime and probably other denominations. A set issued prior to April in 1865 would also have lacked the dime.
The charges of $3, $43 and $46 for "silver," gold and complete proof sets remained unchanged from 1858 through 1867 at least, by which time three new denominations had been added in the minor coins, namely the 2¢, 3¢ nickel and 5¢ nickel. In the Boston Public Library, call number CJ5802. P43, is a copy of the 1867 Mint Regulations or "Circular Letter in Relation to American Medals and Cabinet Coins, July 1, 1867", promulgated by Henry R. Linderman, then Director. Rule 5 in this revision maintained the prices "as heretofore." It would appear that 8¢ of the premium was a proofing charge, the remainder being for wrappers and lagniappe. The unique bound volume Mint Cabinet Accounts and Memoranda, 1857-1904, in custody of Dr. Vladimir Clain-Stefanelli, Curator of Numismatics at the Smithsonian Institution, testified that the curator of the Mint Collection obtained proof sets directly from the Coiner, uncased and probably unwrapped, at face value plus 8¢ specie, higher amounts being paid during the 1867-79 period in the more or less depreciated greenbacks. I shall include the data from this valuable source in the tabulation by dates. The records of bills presented by the first curator of the Mint Cabinet Collections (William Ewing DuBois), 1839-61, naming every purchase or other accession made for the collections of coins, medals, minerals and ore samples, counterfeits, and other oddities, supposedly went with other Mint documents of the period to the National Archives, but in 1953 I could not find any despite intensive search. Should any have survived the selective destruction in 1925, presumably R. W. Julian or some other recent researcher will sooner or later locate them, and from these bills we might learn a great deal about the time of issue of proofs, patterns, and other material, the nature of trades made by the Mint Cabinet officials with various collectors (such as the famous trade of a duplicate 1804 dollar with Matthew Stickney for various Colonials), and other valuable data enabling filling of some of the remaining lacunae in our knowledge of American numismatics.