Recognized HTML document


head, though low, but after all it was not so very large in circumference. Of those persons whom I have mentioned in the foregoing chapters, the heads of William Spottiswoode and Mr. Gassiott were larger round ; Professor Sharpey's was the largest of all. A slight want of symmetry on which Mr. Gladstone laid stress was no peculiarity at all, for the heads of normal persons are rarely quite symmetrical.

The " Measurement of Resemblance " between portraits is a subject on which I have been engaged off and on during late years, and which I hope to take up again. The best of my ideas at present is to prepare a strip of card one inch broad and printed with numerals of various standard sizes from i to 9. Then to mount the portraits on slides actuated by strings, and to station them at such distances that the interval between the pupils of the eyes and the mouth in each portrait shall be apparently the same as the breadth of the strip. Then to interpose a wedge of tinted glass in front of an eye-hole, and to slide it until the portraits become indistinguishable. In that position to read off the smallest of the standard numbers that is simultaneously legible. I have made many experiments, differing in particulars, and described one of them in Nature, October 4, 19o6 [176], which seems to me not so good as the one briefly outlined above.

The chief value to me of the Laboratory during the latter part of the time of its existence, and the reason why I continued it so long, lay in the convenience it afforded for obtaining and testing the value of finger-prints. My interest in them arose through a request to give a Friday evening lecture