"As a large number of persons pass through the Indian gaols not only while undergoing terms of imprisonment, but in default of payment of a fine, it could not but prove of value were the finger prints of one and all secured. They might assist in identifying persons who have formerly been convicted, of whom the local police have no knowledge, and who bear a name that may be the common property of half a hundred in any small town." Whatever difficulty may be felt in the identification of Hindoos, is experienced in at least an equal degree in that of the Chinese residents in our Colonies and Settlements, who to European eyes are still more alike than the Hindoos, and in whose names there is still less variety. I have already referred (p. 26) to Mr. Tabor, of San Francisco, and his proposal in respect to the registration of the Chinese. Remarks showing the need of some satisfactory method of identifying them, have reached me from various sources. The _British North Borneo Herald_, August 1, 1888, that lies before me as I write, alludes to the difficulty of identifying coolies, either by photographs or measurements, as likely to become important in the early future of that country. For purposes of registration, the method of printing to be employed, must be one that gives little trouble on the one hand, and yields the maximum of efficiency for that amount of trouble on the other. Sir W. Herschel impressed simultaneously the fore and middle fingers of the right hand. To impress simultaneously the fore, middle, and ring-fingers of the right hand ought, however, to be better, the trouble being no greater, while three prints are obviously more effective than two, especially for an off-hand comparison. Moreover, the patterns on the ring-finger are much more variable than those on the middle finger. Much as rolled impressions are to be preferred for minute and exhaustive comparisons, they would probably be inconvenient for purposes of registration or attestation. Each finger has to be rolled separately, and each separate rolling takes more time than a dab of all the fingers of one hand simultaneously. Now a dabbed impression of even two fingers is more useful for registration purposes than the rolled impression of one; much more is a dabbed impression of three, especially when the third is the variable ring-finger. Again, in a simultaneous impression, there is no doubt as to the sequence of the finger prints being correct, but there may be some occasional bungling when the fingers are printed separately. * * * * * For most criminal investigations, and for some other purposes also, the question is not the simple one just considered, namely, "Is A the same person, or a different person from B?" but the much more difficult problem of "Who is this unknown person X? Is his name contained in such and such a register?" We will now consider how this question may be answered.