The pattern in every distinct finger print, even though it be only a dabbed impression, contains on a rough average thirty-five different points of reference, in addition to its general peculiarities of outline and core. They consist of forkings, beginnings or ends of ridges, islands, and enclosures. These minute details are by no means peculiar to the pattern itself, but are distributed with almost equal abundance throughout the whole palmar surface. In order to make an exhaustive comparison of two impressions they ought to be photographically enlarged to a size not smaller than those shown in Plate 15. Two negatives of impressions can thus be taken side by side on an ordinary quarter-plate, and any number of photographic prints made from them; but, for still more comfortable working, a further enlargement is desirable, say by the prism, p. 52. Some of the prints may be made on ferro-prussiate paper, as already mentioned pp. 51, 53; they are more convenient by far than prints made by the silver or by the platinum process. Having placed the enlarged prints side by side, two or three conspicuous and convenient points of reference, whether islands, enclosures, or particularly distinct bifurcations, should be identified and marked. By their help, the position of the prints should be readjusted, so that they shall be oriented exactly alike. From each point of reference, in succession, the spines of the ridges are then to be followed with a fine pencil, in the two prints alternately, neatly marking each new point of comparison with a numeral in coloured ink (Plate 13). When both of the prints are good and clear, this is rapidly done; wherever the impressions are faulty, there may be many ambiguities requiring patience to unravel. At first I was timid, and proceeded too hesitatingly when one of the impressions was indistinct, making short alternate traces. Afterwards on gaining confidence, I traced boldly, starting from any well-defined point of reference and not stopping until there were reasonable grounds for hesitation, and found it easy in this way to trace the unions between opposite and incompletely printed ends of ridges, and to disentangle many bad impressions.