Woods: What Is There in Physiognomy? 



317 



Albums" to see if the results would fit 

 with those already obtained, and also 

 to see if there is any marked tendency 

 for the nose to enlarge with age. 

 The data are not sufficient to answer 

 the latter question but it appears that 

 the changes with advancing age are 

 merely in the nature of a greater 

 variability. Probably in some, the 

 nose continues to grow, in others the 

 rest of the face becomes heavy in com- 

 parison with the nose. The Canadian 

 soldiers were comparatively young, 

 probably about the age of college 

 graduates. The Harvard photographs, 

 being from a presumably higher selected 

 social and intellectual group in the 

 community, ought to show a higher 

 percentage of large — long and a lower 

 percentage of short — small than the 

 enlisted men from Canada. 



From the Harvard Class Album of 

 1920 I made the ratios 9—34—19 on 

 the first 60. Mr. Wiggam's inde- 

 pendent estimate was 29 — 21 — 10. 

 The two combined give 38 — 55 — 29. 



I then went through Report VII of 

 the Class of 1886. Here photographs 

 are given of the students at the time of 

 graduation, and also pictures taken 

 twenty-five years later. I examined 

 the first sixty examples, first the young 

 graduates in turn, covering up with a 

 slip of paper, all the adults, and then 

 repeated the process concealing the 

 youthful likenesses and concentrating 

 on the adults. In this way one can 

 measure a double error. If the nose did 

 not change in proportion to the rest 

 of the face, and if one's judgment were 

 perfect, one ought to pick exactly 

 the same individuals for the three 

 classes (r = 1 .00) . Even using the same 

 identical photographs, one's second 

 judgment is not identical with one's 

 first. The correlation is, as above 

 stated, about r = .60 to r = .80. As 

 the correlation which I obtained for 

 the Class of '86 was r = .39 approxi- 

 mately, and that for the Class of '87 

 r = .38 approximately, it would seem 

 that there is a good deal of change in 

 the proportions of the features, but 

 nevertheless there is enough constancy 

 to yield a significant correlation, even 



after allowing for necessary errors in 

 judgment. 



The two squares below show that all 

 figures, from the Harvard tests, fit in 

 about where they should for persons 

 mentally above the level of the "aver- 

 age man" but below the level of the 

 "great man." 



ADULTS 



Aged About 47 

 Large Small 



or Long Average or Short 



15 



Totals 



20 29 



Class of 1886 

 r = .39 approx. 



11 



Large 



or 

 Long 



ADULTS 

 Aged About 47 

 Large Small 



or or 



Long Average Short Totals 



14 



c Average 4 31 5 40 



^ Small 

 or 

 Short 



Totals 12 39 9 



Class of 1887. 

 r = .38 approx. 



HEREDITY AND PHYSIOGNOMY 



All this "counting noses" is not 

 without significance for the science of 

 heredity. Great men, famous men, 

 eminent men, distinguished men, suc- 

 cessful men, high average men, — -using 

 these terms as a descending series of 

 grades towards mediocrity, are all to a 

 certain extent the product of outward 

 circumstances as well as inward forces 

 predetermined in the germ-cells. It 



