It requires considerable patience and caution to arrive at trustworthy conclusions, but it may emphatically be said that there is no _peculiar_ pattern which characterises persons of any of the above races. There is no particular pattern that is special to any one of them, which when met with enables us to assert, or even to suspect, the nationality of the person on whom it appeared. The only differences so far observed, are statistical, and cannot be determined except through patience and caution, and by discussing large groups. I was misled at first by some accidental observations, and as it seemed reasonable to expect to find racial differences in finger marks, the inquiries were continued in varied ways until hard fact had made hope no longer justifiable. After preliminary study, I handed over the collection of racial finger prints to Mr F. Howard Collins, who kindly undertook the labour of tabulating them in many ways, of which it will be only necessary to give an example. Thus, at one time attention was concentrated on a single finger and a single pattern, the most instructive instance being that of arches on the right fore-finger. They admit of being defined with sufficient clearness, having only one doubtful frontier of much importance, namely that at which they begin to break away into nascent-loops, etc. They also occur with considerable frequency on the fore-finger, so the results from a few hundred specimens ought to be fairly trustworthy. It mattered little in the inquiry, at what level the limit was drawn to separate arches from nascent-loops, so long as the same limit was observed in all races alike. Much pains were taken to secure uniformity of treatment, and Mr. Collins selected two limits, the one based on a strict and the other on a somewhat less strict interpretation of the term "arches," but the latter was not so liberal as that which I had used myself in the earlier inquiries (see p. 114). His results showed no great difference in the proportionate frequency of arches in the different races, whichever limit was observed; the following table refers to the more liberal limit:-- TABLE XXX. _Frequency of Arches in the Right Fore-Finger._