Fig. 3.

Dabbed down simultaneously. Dabbed down simultaneously.
Rolled separately. Rolled separately.
Left hand. Right hand.

Form of card used for impressions of the ten digits. 11½ × 5 inches.

Fig. 4.

Roller seen from above. Side view.
The outer ring is to be taken
as representing either a thin
india-rubber tube, or a thick
layer of the composition
used in printers’ rollers.
End view.

Roller and its bearings, of a pocket printing apparatus.

The cores give great assistance in breaking up the very large groups of all-loops (see Table XII., Nos. 11 and VIII.); so does an entry of the approximate number of ridges in some selected fingers, that lie between the core and the upper outline of the loop.

The plan I am now using for keeping finger prints in regular order, is this:—In the principal collection, the prints of each person’s ten digits are taken on the same large card; the four fingers of either hand being dabbed down simultaneously above, and all the ten digits rolled separately below. ([Plate 2], Fig. 3.) Each card has a hole three-eighths of an inch in diameter, punched in the middle near to the bottom edge, and the cards are kept in trays, which they loosely fit, like the card catalogues used in many libraries. Each tray holds easily 500 cards, which are secured by a long stout wire passing like a skewer through the ends of the box and the holes in the cards. The hinder end of the box is sloped, so the cards can be tilted back and easily examined; they can be inserted or removed after withdrawing the wire.

It will be recollected that the leading and therefore the most conspicuous headings in the index refer to the fore, middle, and ring-fingers of the right hand, as entered in column A of the Specimen Register (Table IX.) The variety of these in the “i and o fore-finger” method, of which we are now speaking, cannot exceed thirty-six, there being only four varieties (a, i, o, w) in the fore-finger, and three varieties (a, l, w) in each of the other two; so their maximum number is 4 × 3 × 3 = 36. The actual number of such index-headings in 500 cases, and the number of entries that fell under each, was found to be as follows:—