RESULTS. 27 



One sees that, in general, and excepting the tendency of a light 

 or medium-colored woman to mate with a man of the same grade, 

 the matings are the less frequent the darker the grade of color of the 

 selected male. All of this selection tends toward an increase in the 

 proportion of white and light-colored offspring in successive generations 

 of the offspring of mulattoes. 



X. The Agreement of the Hypothesis with Popular Observation and 



Nomenclature. 



In all countries where extensive hybridization has taken place 

 between black and white a nomenclature has grown up which it is 

 interesting to consider. This nomenclature indicates a rough attempt 

 to express the proportion of Caucasian (or negro) blood as measured 

 by skin color. We may examine this nomenclature to see to what 

 extent it accords with our fundamental hypothesis. 



The word mulatto is widely used to indicate the first generation 

 of hybridization between black and white. It is used loosely to indi- 

 cate any person with a grade of skin like that of a mulatto a grade 

 which, as we have seen, is expressed quantitatively by 26 to 40 per 

 cent of black. The children of two mulatto parents (according to 

 Johnston, 1910, p. 55) are styled cascos in Spanish America, but in 

 the countries where we have studied we have not found a specific 

 term for the produce of this mating. This is the more significant as 

 the mating is fairly common. Failure to apply a term here may well 

 be a consequence of the great variability in this generation. 



To the produce of a mulatto and a pure white is given the term 

 quadroon. This corresponds, on the average, to our one-factor or 

 light-colored, but, as we have seen (table 10), strict quadroons vary 

 from white to medium-colored. This term is also used practically as 

 synonymous with light-colored and does not always imply precise 

 knowledge of ancestry. 



To the produce of a mulatto and a full-blooded negro is applied 

 (in Jamaica and the United States) the term "sambo;" also in Ja- 

 maica, "mangro" is applied to the progeny of a "sambo" and a 

 full-blooded negro. This mating realty gives, by hypothesis, a variable 

 progeny, and the term is practically applied to the 3 -factor or dark- 

 colored condition, without regard to ancestry. Thus for two mulatto 

 parents one-fourth of their children are sambos. 



For social reasons the classification of the matings of quadroons 

 with whites has been carried still further. But here the basis of 

 classification is generally the pedigree rather than the skin color. The 

 social significance of this will be discussed later. Here attention may 

 be called to the nomenclature used in Jamaica, according to which the 

 child of a quadroon and a pure-bred white is an octoroon; the child 

 of an octoroon and a pure Caucasian is a mustifee ; the child of a musti- 



