INHERITANCE OF VARIATES IN THE ARMADILLO 163 



At best an unreal appearance of blending might be produced by 

 the fact that a large region, conceived of as a mosaic of pure 

 maternal or pure paternal territories, might as a whole strike a 

 kind of spurious average between the numbers characteristic of 

 the two parents. As yet, however, I see no necessity of positing 

 any intra-individual segregation more elaborate than that in- 

 volved in the dominance of the paternal influence on one lateral 

 half and that of the maternal on the other half. That the right 

 and left half of the armor element are rather independent regions 

 from the standpoint of their origin is clearly appreciated when we 

 realize that before the invagination and closure of the neural tube 

 the regions of embryonic ectoderm destined to form the right and 

 left sides of the shell are widely separated from each other and, 

 only on the closure of the neural tube and the healing together 

 of the two wings of ectoderm, is a single shield produced. Since 

 unquestionably the hereditary potentialities of these ectodermal 

 regions was settled long before the period of embryonic develop- 

 ment in question, it is not difficult to understand that one part 

 might be in its tendencies quite different from the other; that one 

 might be dominated by the maternal forces and the other by the 

 paternal. 



5. POLYEMBRYONIC AND UNIPARENTAL CORRELATIONS IN INDI- 

 VIDUAL BANDS 



In order further, to test the validity of conclusions expressed in 

 the earlier paper to the effect that, although the total number of 

 scutes in the banded region is very rigidly predetermined, their 

 alignment into bands is a mechanical process not subject to heredi- 

 tary control, I shall herewith present further data derived from 

 the new collection of material. It does not seem worth while to 

 take up the large amount of space that would be needed for tabu- 

 lation of the scute counts by bands of the 115 sets dealt with in 

 table 1. Instead I shall use only 20 male and 20 female sets com- 

 prising a large part of the collection of fetuses marked 'C 

 Forty sets are considered ample for our present purposes for the 

 reason expressed above that even the original twenty sets give 

 results almost identical with those determined for the much larger 

 number dealt with in previous paragraphs. The reader is referred 



