The first perforated sheets of the three cents were from the same plate as the last unperforated sheets, and consequently have the rectangular outside frame lines, not only at the sides, but at the top and bottom as well. As the stamps measure 25 mm. vertically and are only 1 mm. apart, and the horizontal rows of perforation are about 25½ mm. from center to center of the holes vertically, the perforations generally cut into the stamp and partially obliterate these lines. As the stamps are only 1 mm. apart at the sides, and are 19½ mm. wide, and the vertical rows of perforations are 20½ mm. from center to center of the holes horizontally and the holes are nearly 1 mm. in diameter, the side perforations also usually cut into some part of the stamp. It is therefore quite difficult to find good specimens of this variety, and to distinguish some of the minor varieties, as the corners are generally imperfect. Specimens were found showing the sheet cut along the colored vertical line, and (X) perforated between this line and the stamp, from either half of the sheet. New plates were however, soon constructed. In one of these, No. 24, the side lines are drawn on the plate from the top to the bottom, and are about 19½ mm. apart. The fine outer lines at the top and bottom are entirely omitted. The maker's imprint, "Toppan, Carpenter & Co., Bank Note Engravers, Phila., New York, Boston and Cincinnati," 1¼ mm. from the outer rows, is 68 mm. long. "No. 24 P," is 4 mm. from the outer rows. The vertical center line is 1¾ mm. from the stamp. The sheet measures 418 mm. from side to side, and 252½ mm. from top to bottom of the printed part. The paper is 447 by 283 mm. The vertical rows of stamps are 1½ mm. apart, and the vertical rows of perforations nearly 21 mm. apart horizontally from center to center of the holes. The last two rows at the sides are a little further apart. The horizontal rows of holes are 25½ mm. apart vertically. Most of the differences in the stamps on this sheet arise from the fact that the central portion is not always placed in the same position in regard to the vertical lines. A few of the stamps show dots in or near the corners. In some the lines are too near some of the corners, in some too far off, and in others they touch and even cut into the blocks. Some few show double or partly double lines. The whole sheet from plate 24, above mentioned, does not contain all the varieties round, nor are they arranged just in the same order that they appear in portions of other sheets examined. All the varieties possible, considering merely the position of the corners and side lines, would be 246. So that each stamp on a sheet might be different in this respect without showing them all. In sheet 24 however, only 32 exist. There are therefore, a number of each variety, as follows, by the table previously given: