Development of the Bureau Precancel

By H. M. Southgate

Although the Post Office Department had authorized the use of precancels since 1887, it was thirty years before the Bureau of Engraving and Printing supplied such stamps to the Department.

Between 1911 and 1915 the Post Office raised with the Bureau the question of competing with local printers for overprinting precancels but apparently the latter’s costs were too high. However, in December, 1915, Director Ralph went into the problem actively and quoted a price of $2.85 per 1000 sheets (of 100 stamps each) for precanceling stamps for 32 of the largest cities.

This proposal was probably based on the supply of stamps showing the name of the city cut into the stamp plate itself, so that the precancelation would be part of the design of the stamp.

The Bureau made up a die, 627- 1-13-16, for a 2c stamp, proof of which was submitted to the Post Office on 1-20-16. At the Post Office’s request further samples showing heavier lines and letters were provided on Sept. 25,1916, as well as a second design which showed the lines and letters in white instead of red. This latter design was made from die 635.

The idea of entering the town name on the plate for individual cities may have been a hangover from the proposal made in 1906. When awards for the 1906 supply of stamps were made, the specifications included a condition that all ordinary stamps above the 2c denomination should be delivered to the post offices of the 26 largest cities and should have,

“Engraved across the face their names and locations, while the others shall be overprinted, from electrotype plates across the face, with the names of the Post Offices respectively and the abbreviated names of the states in which they were located.”

The purpose of the proposal was to prevent robberies. The plan died a natural death but King and Johl (from whose valuable work, Vol. I, p. 18, the above is taken) say: “One die had been made with the Chicago, inscription, but it was never used.” An inspection of the die list does not reveal any such die for Chicago, although it may not be listed as such. However, this idea may well have suggested the design for the precancel shown in Figures 1 and 2.

On 9-25-16, Director Ralph likewise revised his price of Dec. 15, 1915, for overprinting as that was made on the basis of precanceling panes of 100 stamps at a time. By precanceling full sheets of 400 stamps the price can be reduced, he says, to $1.50 per 1000 sheets of 100. The Director further states that he would not care to undertake the work for more than 20 of the larger post offices as each office would require 400 separate electrotypes which would be combined into one form, and these 20 offices would require 8000 of these pieces. This would require a great deal of space for their proper care aside from the cost. It extended to a great number of offices the cost and care of these electrotypes would make the project impracticable. The Bureau was willing to provide precanceled stamps for 20 offices.

The Bureau’s revised bid of $1.50 per 1000 sheets brought their figure down to the local prices for New Orleans, La., Augusta, Me., and Springfield, Mass., which were $1.50, $1.70 and $1.85 per 1000 respectively, and the Post Office Department awarded a contract to the Bureau which was to be in the nature of an experimental run for the purpose of determining costs, and on Nov. 25, 1916, Postmaster General A. S. Burleson advised the Director of the Bureau of Engraving and Printing that the samples submitted on Sept. 25 had been carefully considered but it was concluded that the cancelation cut in the design (figures 1 and 2) on any of the 4 samples was not sufficiently strong to distinguish any of them at a glance and that for the present the idea of printing precanceled stamps from special plates with the precancelation cut into the design will not be pressed. This would seem to indicate that the idea of making the precancelation part of the design was initiated by the Post Office.

December 22, 1916, Postmaster General Burleson advised Secretary W. G. McAdoo of the acceptance of the offer of his Director of the Bureau of Engraving and Printing, J. E. Ralph, for precanceling 4,000,000 stamps each for Augusta, Me., New Orleans, La., and Springfield, Mass., and the Bureau was in the precanceling business for the time being and the “experimentals” were underway.

In the American Philatelist for March, 1917, Joseph B. Leavy, the Government philatelist, gives notice of this innovation but it created little interest and the existence of these Bureau precancel classics was generally overlooked until Frederick W. Allen unearthed the Leavy article and reprinted it in part in Mekeel’s Weekly Stamp News of January 25, 1926.

The design of overprint employed for these experimentals evidently followed that of precancels then in general use. The same style of type was employed for all three cities but where necessary, condensed. The heavy, continuous single line above and below the town designation was employed and has since been a characteristic of all Bureau precancels.

Figure 3 illustrates the experimental type of overprint. The Leavy article advises that the Bureau prepared type set forms of twenty-five subjects in blocks of five by five. The Government Printing Office prepared electrotypes from these forms which the Bureau set up in frames to precancel full sheets of 400 stamps. In order to obtain more exact registration of the overprint on the sheet the 25 subject electros were cut by the Bureau, horizontally into a five by two and a five by three block.

Leavy suggests that this division of the twenty-five subject electro was to meet the shrinkage of the paper in the sheet, but unless the stamps were printed with the grain of the paper across the long dimension of the stamp, which was not the practice at the time, at any rate when the “star” plates were made, there would appear to be no virtue in splitting the blocks horizon tally. Also the split was used to in crease the spacing between rows from 6 mm. to around 8 mm., at least on the New Orleans overprint, so that the split of the twenty-five subject block would appear to be to improve the original setup, and give more accurate vertical registration of the overprint.

It was recognized that this initial order was in the nature of an experiment. The Bureau had been able to compete successfully with only three out of 32 of the largest users of precancels. As no further orders were placed for Bureau production at the time, it would seem that the cost of doing the work was too high to warrant further experimentation along the lines used for local over-printing. Some different method of production would be required.

“Necessity is the mother of invention”, and rather peculiarly the next step in the development of the Bureau precancel was the result of the study by a joint committee of Post Office and Bureau officials appointed for the purpose of examining the methods and details for the manufacture of postage stamps by the use of the rotary press in the Bureau of Engraving and Printing with a view of determining of the advisability of continuing the production of postage stamps by this process. Postmaster General Work on Aug. 4, 1922, appointed a committee consisting of M. J. Eidness, Supt. of the Stamp Division, Richard Breaden, Asst. Supt., and B. L. Andrews, Asst. Supt. Mechanical Division, all of the Post Office Department, and John Hardie, Supt. of Plate Printing, and E. P. Ganger, Supt. Engineering Division of the Bureau.

The Post Office Department at this time was receiving many complaints from the post offices and the public on account of the curling of the rotary press product due to excess gum and insufficient margins making overprinting practically impossible.

The Bureau had recently purchased four large sized rotary presses to handle 400 subject work but so unsatisfactory were the existing stamps from rotary presses except for coil work, that in spite of the great saving effected by the rotary presses, the Post Office had suggested July 19, 1922, that the order be canceled and that the existing large presses be devoted to coil work as the Bureau was unable to build up a coil stock sufficient to meet the governing demands for coils.

The excess gum difficulty could be corrected but apparently the complaint of precanceling trouble was so serious that the committee was ready to stop further rotary sheet productions, until Richard Breaden suggested the idea of precanceling on the rotary presses and supplying the larger offices with the Bureau made precancels.

The first scheme was to precancel stamps in the full roll, running the web of printed stamps through a second rotary press fitted with precanceling plates in place of the stamp plates. To investigate this idea, the Bureau prepared a die shown by Fig. 4. This die, No. 709, is dated Aug. 18, 1922, and rolls 1183 and 1189 were made for the preparation of plates Nos. 14047 and 14046. These plates were never certified apparently, although work was evidently started as they were canceled Nov. 5, 1924. These 400 subject plates were of course of the same dimension as the similar stamp plate. Evidently while the Bureau was pushing the preparation of the plates, the question of cost of over-printing was being worked out. As a second run of the roll through the press would materially increase the cost of production, it might make the price prohibitive.

Then Breaden, with McKenzie of the Bureau, developed the present system of coupling electrotype precanceling plates to the Stickney press so that the overprinting would be done during the same operation as the stamp printing and gumming. There is apparently no record of any experimental work in connection with the development of these electrotype plates. Undoubtedly the Government Post Office conducted numerous trials before a satisfactory curved overprinting plate was made. The small diameter of the roll on which the plates were mounted made the production of satisfactory plates difficult until the technique had been established.

There is no need of repeating the description of the Breaden-McKenzie system in detail. The first overprinting plates were electros made from a hand set, Bureau-made type page and sent to the Government Printing Office for the manufacturing of the electrotype. The G. P.O. makes an electrotype pattern from which model the curved plates are made as ordered by the Bureau. This was the procedure followed in making the blocks for experimentals up to the point of curving. However, instead of mounting the electrotype shell on a wooden backing as the practice for a flat bed press, the curved electros were backed up with a type metal composition to a thickness of about two-tenths of one inch, curved and fitted for direct mounting on the Stickney rotary press.

The first curved precanceling plates were made for New York and the initial shipment consisted of the delivery of 20,000,000 1c stamps on May 2, 1923. Philip Ward in Mekeel’s of May 19, 1923, advises that the first shipment has gone to New York.

The success of the process was such that on August 20, 1923, the Third Assistant Postmaster General was able to advise that a total of 186,200,000 1c stamps (precanceled) had been or would be shipped, by August 21st, to New York, Chicago, Cleveland, Boston, Kansas City, St. Louis, Philadelphia, and Cincinnati. Also, that plates had been ordered for Nashville, Detroit, Minneapolis, Springfield, Ohio, Milwaukee, Atlanta, San Francisco, Buffalo, Dayton, Pittsburgh and Columbus. All of this precanceling was of 400 subject sheets, but in November, 1923, the New York office was advised that as soon as the necessary experimentation was completed, coil stamps would be furnished. The first shipment of 1c coils was made to New York on Jan. 17, 1924, followed very shortly by 1c and 2c coils to Chicago and Boston and 1c coils to Washington.

In order to meet the Christmas precanceling requirements the P.O. asked the Bureau, October 25, 1924, to prepare rotary plates of the 3, 4, 5, and l0c denomination, later dropping the 3c. All denominations under the 11c, except the ½c sheet and 6c coil, were available in precancelled form in both sheet and coil form by the middle of 1927. April, 1929, showed the first of the 1/2c Bureaus while the 6c coils did not appear until August, 1932, to be followed shortly by all values, 11c to and including the 50c.

With the successful production of precanceled coils the committee appointed by Postmaster General Work had solved the main problem caused by the use of the rotary press and the four presses which had been delivered, were taken out of storage and soon put to work, grinding out the Bureau precancels.

Figure 5 shows the original types of overprint used. These are generally known to collectors as “Old Type” precancels. This design was in general use through the perf. 10 and perf. 11×10 1/2 series, to October, 1927, although a few old type plates were employed as late as the spring of 1933.

The Bureau engineers had from the start of production of precancels by the rotary press, recognized that the life of the electrotype plates would be comparatively short and that a tougher plate would be desirable.

Apparently with this end in view consideration was given to the manufacture of steel plates duplicating in general the electrotypes. The Bureau prepared dies 716 and 717 (Figure 6 and 7) dated May 4, 1923 and May 7, 1923, respectively.

Apparently the white letter design of die 716 was considered unsatisfactory us the fine lines might not show up well or perhaps would be liable to break under the pressures required. Possibly they would cut the paper. In any event, there seems to be no record of any transfer roll or plate being made. The size of the lettering would seem to indicate that the design was intended to be photographically reduced for the manufacture of a surface plate.

Evidently the solid letter design was considered preferable as the record shows that roll 1260 was made from die 717 (Figure 7) and experimental plates 14634 and 14635 were started. These 50 subject plates were apparently never certified but were canceled Nov. 5, 1924. Four plates were designed each 8.375 inches long to fit a cylinder 3¼ inches in diameter.

The original system, employing electrotype plates seem to have stood up sufficiently well to make production successful and the original Breaden-McKenzie development is still in use.

In October, 1927, the Post Office adopted a uniform type Overprint for all Bureau precancels in place of the Old Type (Figure 5). This is known to collectors as the standard type or B11 and B13, depending on whether the Overprint is in two or three lines. The patterns for the electrotypes were still made from a hand-set “type page.”

In February, 1932, the font of type used for setting up the “type page” was becoming worn and the substitution of machine set slugs in place of hand set type was adopted. The cost of a pattern made with slugs would be reduced and the liability of error from the use of a wrong letter in a subject avoided. The Government Printing Office could not duplicate exactly the Standard overprint type and the machine set slugs give a slightly different precancelation as shown by Figures 9 and 10. All towns ordering Bureau precancels for the first time since February, 1932, have the machine-set patterns. To date there are five towns which have used both the hand-set and machine-set type for either sheet or coil product, but not both.

Figure 9 shows an enlargement of the hand-set and machine-set (Ludlow) type alphabets almost 4 to 1. Figure 10 shows the actual size of the hand-set and machine-set type. In both cases the hand-set illustration was made from worn type which doubtless accentuates somewhat the boldness of the face.

The development of the Bureau precancels has been remarkably free from changes. This is doubtless one reason for their popularity. There is enough field for collecting in the quest of stamps from the towns using these stamps without added complications that might have been injected by various experiments. Bureau precancels, with exception of the Experimentals are only found on the 4th Bureau series i. e., the 1923 perf. 10 and 1926-1931 perf. 11×10½ issues. The old type overprint appears with both perforations. The standard type is only found on the compound. I am indeed indebted to Messrs. Eidness, Breaden and King at the time or still with the Post Office Department for the background of these notes. Thanks are due to Director Hall and Messrs. Weeks and Benzing for their help with the Bureau end of the story.

 

 

 

 

This article is reprinted from the Sixth Annual Bureau Issues Association Special Edition of the Weekly Philatelic Gossip of May 9, 1936. Copyright, 1936 by The Gossip Printery.