In order to ascertain the mean gauge of the B machine, the only possible way would be to get a line of perforations representing the whole length of the machine, and measure it. The longest line of perforations we have been able to experiment on is one of 124 mm.; this contains 72 holes, giving a mean gauge of about 11⅔ in the 2 centimetres. In some places in this line nearly 13 holes can be counted in the space of 2 centimetres, in others not more than 11, and all intermediate gauges as well are present in the same line. Indeed one has only to move the perforation-gauge one or perhaps two holes to the right or left to obtain a striking change of gauge. We have seen that the same sort of thing occurs in the perforation done by the A machine, but in a much lesser degree, the variation being only between 14 and 15—here it is from 11 to nearly 13, and is visible at a glance without the aid of the perforation-gauge. The above mean gauge of 11⅔ is very near to the 11½ usually ascribed in catalogues to stamps perforated by the B machine, but even if that be correct as a mean gauge for the whole line, it is very misleading so to call the perforation, as a collector whose solitary specimen might gauge, say 13 x 12½, would naturally suppose that it was a variety differing essentially from those said to gauge 11½. This, as we have seen, it would not be. As in the case of the A machine, after the B machine had been in use some considerable time (about 1876), its perforations show a change of character, the holes being seldom punched right through; the discs of paper remain in their places, so that when the stamps are severed the edges are very ragged. Otherwise there is no change whatever, the holes, or the marks where they should be, being still circular, and spaced in the same irregular manner. Besides the two simple perforations A and B, we find one compound when the two machines are used in conjunction for the same sheet. Whenever this compound appears it is invariably the same in all cases; that is, the horizontal lines of perforation are made by the B machine, and the vertical lines by the A machine; or, adopting the philatelic notation now generally accepted, it is “B×A.” Omitting the Six Pence of 1862, this reduces the possible number of varieties in the perforations of the stamps of Section I. to three in all; and in order to show at a glance how these occur in the different issues, we have arranged them in a table, which will be found under the head of Appendix B.