132 THE BIOLOGY OF TWINS 



brought out for the numbers of scutes in the banded 

 region are practically duplicated by those derived 

 from a study of any other part of the armor. Results 

 of this sort would" in themselves be,suf!i.cient to prove 

 the polyembryonic character of armadillo quadruplets; 

 if these individuals were from four germ-cells there 

 would be no reason to expect correlations higher than 

 those that obtain between brothers, which are never 

 much higher than 0.5. From the genetic standpoint 

 a set of armadillo quadruplets is essentially one individ- 

 ual in four parts. AAThen we are comparing one fetus 

 of a set with another, we are dealing with a special 

 case of intra-individual correlation. The proof of this 

 assertion lies in the fact that the degree of correlation 

 is of the same order as those determined for equivalent 

 parts of one individual and is never paralleled by that 

 determined between separate individuals. 



In closing this very brief statement of some of the 

 significant facts about the inheritance of aggregates of 

 integral variates, two points should be emphasized. 

 First, it should be said that these scute number dif- 

 ferences are about the only inherited dififerences found 

 in the species that are available for biometric study. 

 Armadillos are practically all alike in color, size for 

 age, and proportions of body parts. Males and females 

 are also alike except for the genitalia. In view of these 

 circumstances the reader will readily appreciate my 

 choice of characters for the study of heredity. What- 

 ever facts about heredity are to be discovered for this 

 species must, therefore, be discovered in connection 

 with these scutes of the armor. The second point to 

 remember concerning the armor characters is that they 



