RELATIVE POTENCY OF HEREDITY AND ENVIRONMENT 537 



the right animal for this purpose — the nine-banded armadillo of Texas. 

 This armored mammal, belonging to an archaic race, abounds in south- 

 west Texas. We discovered that it always produces not merely twins 

 but quadruplets at every pregnancy and always only one egg is fertilized 

 at each pregnancy. Over two hundred litters sufficiently advanced for 

 detailed comparisons were examined, and in this material we have the 

 data for determining just exactly what are the extreme Hmits of hered- 

 ity and how much leeway there may be for environmental modification. 

 The animals are beautifully designed for detailed comparison. Their 

 nine bands of armor, composed of definite scutes, the rings of scutes on 

 the tail, the definite scale pattern of the head, all these and many other 

 characters lend themselves to exact statistical treatment. 

 Some of the results of such a comparison are these: 



a) All individuals of a given litter, derived from one egg, are of 

 the same sex. There are no exceptions. 



b) The coefficient of correlation of a considerable number of body 

 characters for 115 sets of quadruplets which showed no marked anom- 

 alies of scute pattern was approximately .93, where i is complete 

 correlation or exact correspondence. This may be taken to prove that 

 heredity accounts for 93 per cent of the bodily characteristics and other 

 factors for about 7 per cent. 



c) WTien, however, individual sets of offspring were examined, very 

 great differences revealed themselves. With regard to certain char- 

 acters some sets showed .99+ resemblance and others showed very Ut- 

 tle more resemblance than would be expected in unrelated individuals. 

 Yet every set of quadruplets is beyond question the product of the 

 division of a single fertilized egg. 



d) In some of these sets of quadruplets where the differences 

 among individuals were great, it was found that one, two, or three 

 individuals closely resembled the mother, while the others were quite 

 unlike the mother, and presumably like the unknown father. 



e) In several sets where the mother had some rare pecuUar arrange- 

 ment of the scutes, one, two, or three offspring repeated the maternal 

 character more or less definitely, while the others failed to show any 

 trace of such peculiarity. 



/) Frequently an asymmetrical peculiarity inherited from the 

 mother was found on the left side of one twin, the right side of another, 

 sometimes on both sides of a third, and wanting in the fourth. 



All of these results indicate that there is operating here a third 

 factor, neither heredity nor environment, but what we might call a 



