464 CHAMBERLAIN. 



235 mixed types. Eeducing these to the percentage basis we have, for 

 each 100 blonds, 79 brunettes and 126 of the mixed type. It will be 

 observed that the proportion of the latter type is lower than the Army 

 average (Table XXIV), but that the proportion of blonds as compared 

 with brunettes is higher than the average for enlisted men of the Army 

 in the period from 1909 to 1911 (100 to 79 as contrasted with 100 

 to 86). 



These facts tend to show that in an average tropical service of 5.5 

 years the relative resistance of the blonds is not so impaired as to lead 

 to a depletion of their ranlfs disproportionate to that occurring among 

 the brxmettes. To refute these deductions it might be argued that the 

 proportions of the three complexion types among the American troops 

 may not have been the same 6, 8, 10 or 12 years ago, at which times 

 many members of the S-C-P-Grroup severed their connections with the 

 regular Army, and that originally the blonds may have been much in 

 excess of the brunettes in the S-C-P-G-roup. We have no statistics as 

 to the proportions of the different complexion types in the Army for 

 the period 1898 to 1901 and it is possible that in the earlier part of this 

 period blonds may have been relatively somewhat more numerous, es- 

 pecially among the Western volunteers. However, we do have some 

 statistics gathered in 1902 and 1903 by Woodruff, (2) and tabulated on 

 page 215 of his book on Tropical Light.^^ Out of 1,294 soldiers there 

 were 420 blonds, 356 brunettes, and 518 mixed types. Eeduced to the 

 percentage basis he found for each 100 blonds, 85 brunettes and 123 of 

 the mixed type. On comparing these figures with those in the last line 

 of Table XXIV it is seen that the blonds were not appreciably more 

 numerous in the Army in 1902-3 than they were in 1909-11, and that 

 they were considerably less numerous than they are at the present time 

 (1911) in the S-C-P-Group. 



SUMMARY OF PAET V. 



The ratio of invaliding home is lower for thp blonds than for the 

 brunettes. The ranks of the blonds in the S-C-P-Group after 6.5 years' 

 tropical service, have not been depleted, the proportion of the fair com- 

 plexion type being probably higher in the S-C-P-Group now than it 

 was among the class from which that group was originally drawn. 

 These are strong arguments in favor of the assumption that the blonds 

 are quite as well able as the brunettes to resist the tropical conditions 

 met with in the Philippines at the present time. 



" Our system of classifying men into blonds and brunettes is the same as 

 that described by Woodruff. His ratio of blonds to brunettes agrees closely 

 with ours, but we find a considerably larger mixed class than he reports. 



