each set of prints may be simply numbered, and placed consecutively in shallow drawers, with 50-100 in a drawer, while the classification may be made by means of a card catalogue, which would contain a numbered card for every individual, arranged in accordance with the system just described. I have arranged my own print collection upon this latter system, the prints being numbered and arranged consecutively in the order taken, while the corresponding cards are classified in accordance with the formulæ, those of the left hands taken first, and subdivided by those of the right.
As for the space required for a collection. If the prints are filed in shallow drawers or slides holding 50 sets each, a collection of 100,000 sets could be accommodated in 2,000 drawers; and allowing a frontage of 16 x 3 inches for each drawer with its surroundings, these would make a cabinet 6 feet high and composed of 56 perpendicular stacks which together would occupy a wall length of 75 feet. An 18-drawer card index with a capacity of more than 20,000 3 x 5-inch cards, as taken from a recent catalogue of office furniture, is 44 inches wide and 14 in height, and the five necessary to accompany the collection in question would form a single stack 6 feet high and 44 inches wide, which will add approximately four feet to the seventy-five given above. Thus a room having 80 feet of wall, linear measure, or a smaller one with a double stack running through the center, would be amply sufficient for the entire collection, properly arranged.
This calculation appears to answer in the affirmative the question of the practicality of keeping palm and sole records of all citizens as advocated in my previous article. M. Bertillon has pointed out the numerous cases in civil life in which one's identity is in peril, and looks forward to a future in which some record, based upon physical characters, will be made of every citizen, but the trouble and inconvenience attending his own system of measurements and the fact that they are applicable during adult life alone, would leave them hardly to be considered for such a purpose.
The palm and sole system, which I originally presented as an extension of the system of Mr. Galton, appears to supply the need in this respect, as the records are easily taken, unchanged from birth to death, quickly compared, either with the hands and feet themselves or with other prints, and capable of brief characterization and of accurate classification by means of simple formulae, to all of which may be added, as their most important advantage, that of absolute certainty, while the Bertillon measurements afford no more than a strong probability.
Prints could be taken in each township or municipality, and filed away in any convenient spot, perhaps the court house of each county
