deduction is, then, exceedingly scanty—if, indeed it can be called evidence. Nevertheless, I think his main conclusion holds good. Independently of his reasoning I had come to exactly the same result in a purely inductive way." He then quotes a number of travelers to the effect that marriage between members of different races produce a phenomenal excess of female births. When we consider the extraordinary proficiency in fiction attained by many travelers in strange lands, we are forced to the belief that Westermarck based his own conclusion on still more scanty evidence.

The statistics given by Dr. Düsing for Prussia^[32] are as follows:

and for mixed marriages:

In the face of these statistics it is impossible to deny that endogamy within a great social class or an ethnic race may have some tendency to produce an excess of male births, while exogamy in this broad sense may diminish the masculinity. But the perpetuation of a comparatively pure race by marriage within that race, and consanguineous marriage in the narrower sense are different propositions. It may easily be that the marriage of individuals of a similar type regardless of consanguinity produces a greater excess of male offspring. According to the percentage of first cousin marriages among the Jews as given by Mulhall,^[33] and allowing the average number of children to a marriage, there would be only 3100 children of such marriages among the Jewish births in Prussia, and in order that these might raise the masculinity of Jewish births even from 106 to 107 the 3100 births would have to have a masculinity of 200. Among Protestants, or especially among Catholics where the percentage of cousin marriage is much smaller, it seems hardly reasonable that the general masculinity would be appreciably affected. A much better case can be made for similarity or difference of race as the cause of the variation. The difference between Catholic and Protestant is, roughly speaking, the difference between the brachycephalic brunette Alpine race and the dolichocephalic blonde Baltic race. So that a mixed marriage in Germany would almost always mean the crossing of two distinct types.

The investigations of M. Gache in Buenos Ayres covering the period from 1884 to 1894 inclusive, show that cross breeding has had the effect of raising the masculinity. The births resulting from unions of Italian, Spanish and French male immigrants with native-born Argentine females, show a higher masculinity than the births produced either by pure Argentine alliances or by pure alliances of any of these nationalities of Buenos Ayres. Further, the unions of Argentine males with females of foreign nationality provide a higher masculinity than is common among Argentines themselves.^[34] These facts do not necessarily contradict the theory that any crossing of great racial groups diminishes masculinity, for all of the nationalities involved in this study are predominantly Mediterranean in blood. The theory is borne out by the statistics of the negroes in the United States, a large proportion of whom are of mixed blood. For taking as a basis the number of children of negro descent born during the year ending June 1, 1900 reported by the Twelfth Census, the females predominated, giving a negative masculinity of 99.8. Furthermore, the percentage of consanguineous marriage is probably high in the colored population.

The following table compiled from Mulhall^[35] and other sources fails to show any correspondence between the percentage of first cousin marriage and the masculinity:

It is impossible to obtain the actual masculinity ratio for the United States, for the Census gives the statistics for only one year in ten and even then is untrustworthy on this point. In a few states birth registration is attempted but the figures thus obtained do not harmonize with the Census and the situation is not greatly improved.^[38] The masculinity varies considerably in different parts of the country, and is generally higher in states where the rural population predominates. This fact agrees with European statistics which almost universally show a high masculinity in rural districts. Table XII, illustrates this point: