A rather interesting study, particularly for the collector of entires, is that of postmarks and cancellations, and sometimes much assistance in the solution of knotty questions is rendered by these often despised and neglected adjuncts to the proper use of postage stamps.

The early cancellations of Canada have been the subject of some attention, more so, in fact, than the postmarks, as they were required to be used on the stamps while the postmark was struck on the cover, where the date and place of mailing would be plainly visible. In one of the early volumes of reports it is stated that "Office Stamps and Seals were supplied from England on 21st July, 1851." It is presumed that this included postmarks and cancellations.

It will be remembered, perhaps, that in the circular announcing the issue of stamps in 1851[77] it was ordered that "Stamps so affixed are to be immediately cancelled ... with an instrument to be furnished for that purpose." The first one so supplied was the "concentric rings" cancellation, consisting of seven concentric circles and having an outer diameter of 18 mm. This is the most common of all, being found from the very earliest dates down to 1870, at least, as it occurs on the early shades of the "small" cents issue. It was generally struck in black ink, but may occasionally be found in a dull blue. A good illustration of this cancellation is seen on the cover numbered 90 on [Plate VI].

By 1855, at least, a modified form of the concentric ring cancellation was introduced. This had a number in the center in large figures, some 8 mm. high, with four concentric circles enclosing it, the outside diameter being about 23 mm. This type was generally struck in black, but is sometimes found in a dull blue also. It can be seen on the strip of stamps numbered 81 on Plate V. The numbers, of course, were placed in the cancellations with a definite purpose, and a little study of entire covers shows that certain numbers were

assigned to certain post offices, as might be suspected. Number 21, for instance, is the most common one and will be found to be connected with Montreal. Further study will reveal the fact that the names of the post offices were taken in alphabetical order, and the numbers assigned to them consecutively in that way. Still further inspection develops the fact that most of the post offices were those in Upper Canada (or Canada West), while but a few of the most important ones were included from Lower Canada (or Canada East.)

Mr. Edgar Nelton seems first to have made a study of these numbers in an attempt to identify their corresponding post offices, and he published a list of some twenty-two as the result of his examination of many original covers.[78] The numbers run up to 52 at least, and using the facts that we have deduced concerning the arrangement of the names, we have endeavored to fill out his skeleton list with such offices as it seems possible may yet be identified with the corresponding numbers. We have done this with some assurance for the following reasons:—

We were fortunately able to examine a Canada Directory for 1857-8, and on looking up the postal information given therein, found a list of the money order offices then existing. This was in two sections, the first containing the names, alphabetically arranged, of 31 offices in "Class No. 1," which included most of the principal cities and towns; and the second a lengthy alphabetical list of offices in "Class No. 2." The first section had a somewhat familiar appearance, and inspection showed that a majority of the names on Mr. Nelton's list of numbered cancellations were there in proper order! But 21 more names were needed, according to the cancellation numbers, to fill out the latter series. The second section was therefore examined for such towns as had the largest populations and were presumably most important. The result enabled more than one name, already on Mr. Nelton's list, to be fitted in its proper place! Here, then, was apparently the solution of the first series of numbered cancellations, and we hazard a guess that the 52 names are the original list of money order offices, arranged when the money order system was instituted in February, 1855.

The subjoined table gives the list of post offices and their corresponding numbers, which has been worked out along the lines above mentioned. It is offered in the hope that more will be done to determine positively the cor

respondence between the two. The names in ordinary type are those that have been identified without any reasonable doubt; those that have been fitted in tentatively are in italics. The Roman numeral following indicates the Class to which the Money Order Office belongs.

LIST OF NUMBERED CANCELLATIONS.