1886.J ^O [Frazer. 



Composite Photography applied to Handwriting. By Dr. Persifor Frazer. 



{Read before the American Philosophical Society, January 16, 1SS6.) 



The following preliminary note on this subject appeared in the Journal 

 of the Franklin Institute for February, 1886 : 



Francis Galton was the first to point out in fugitive memoirs, and nota- 

 bly in his important work, "The Human Faculty," that one could sift 

 the common from the accidental features of a number of objects by 

 exposing them in succession to a sensitized plate in such a manner 

 that the images of the similar parts of the different objects should 

 occupy as nearly as possible the same parts of the plate ; and that 

 each object should be exposed for only a fraction of the length of 

 time necessary to complete a picture on the film used. This fraction 

 depended generally, if not always, on the number of objects and on 

 the sensitiveness of the film. For example, if there were eighteen ob- 

 jects and the plate took thirty-six seconds to develop, each object would 

 ordinarily be exposed for two seconds. It is easy to see that the result in 

 the finished picture would be that those features which all the objects had 

 in common would be re-enforced by each separate exposure, whereas those 

 features which were accidental or variable, and which would be different 

 for every individual, would be exposed for but two seconds and would be 

 so indistinct as practically to fade away. Where the object was to catch a 

 family likeness by exposing all the members male and female to the same 

 portion of the plate, the result xs a curious medley of faint whiskers and 

 moustache ; of hair parted in the middle and at the side ; of female gowns 

 with buttons to the throat and of male shooting jackets thrown open. But 

 out of all this faint halo of confusion and blur, there starts a characteristic 

 face which is the family type. Very often, too, this type-face resembles no- 

 ticeably two different members of a family between whom no one can find 

 a resemblance. It is this latter fact (which might have been expected) that 

 induced me to look to the process for aid in solving the problem of 

 identity of origin in handwriting. When a number of animals of the same 

 race are thus treated, the method secures the fixing of the race or family 

 characteristics, etc., as the case may be. When a number of pictures or 

 coins bearing different representations of the same individual or scene are 

 the objects, the result is to obtain either the average appearance of the same 

 thing under different conditions (as for instance a man at different times 

 of life), or the average of the impression made by identically the same 

 thing on different artists. In this case, the merit of the process is that it 

 constructs its image out of all that many pairs of trained eyes have seen, 

 without giving undue weight to auy one pair. So far, then, these efforts 

 have been directed to re-finding a lost or concealed existence through mul- 

 tiple testimony, very much as the law tries to get at the truth by examin- 

 ing a number of witnesses. 



At first sight one would suppose, however, that the case of handwriting 



PROC. AMER. PHILOS. SOC. XXIII. 123. 3C. PRINTED AUG. 7, 1886. 



