Recognized HTML document

164   Life and Letters of Francis Galton These are represented in the following diagram:

4

wsv

. x{11

well

Sw

By

BW

wsv

sv

Sw

W Sv

Sw

BV

BV

WBV

BW

Bw

W BV

Fig. 22. Classifying by nature of Ridges from Deltas.

We can, perhaps, improve somewhat Galton's indexing in the following inanner*. Consider the digits taken in order from little finger of left hand to little finger of right (as the hands are placed palms downward on the knees) to occupy the places from first to last of a ten-figure number, e.g. 32881,56490, then this would be interpreted as meaning that the little finger of the left hand was SV - B W, the ring finger SW- B V, the pointer and the forefinger WSV - B W and the thumb WSV - WB V; the thumb of the right hand would be WS V - B V, and so on down to the little finger of the right which would be an arch. Thus a thousand million variations would be possible, and every individual would have his own ten-figure index number, which could be

Fig. 23. "Outlining" a rolled pattern.

recorded in numerical order in the index. The question, however, of how many of these would be "repeats" remains to be considered. Galton shows how, after outlining the pattern, it is fairly easy to classify a great variety of patterns according to his scheme (see his Fig. 9, p. 7, which contains forty

* Galton later drops without comment his classification of prints from the contours of the cores. He nowhere, states why. Probably he found it not adequately discriminative for large numbers, or perhaps he discovered the personal equation involved in drawing contours.