174 H. H. NEWMAN 



It will be shown in this section of the paper that, although in 

 one region of the armor, for example the banded region, the ma- 

 ternal scute number may be dominant, other large sections of the 

 armor may show a dominance of the paternal scute number. In 

 brief there may be a mosaic of maternal and paternal regions of 

 dominance involving larger regions than any that have been 

 dealt with in the earlier parts of the paper. The data upon which 

 this study is based is given in full in tables 5, A and B. 



1. POLYEMBRYONIC CORRELATIONS FOR THE FIVE MAIN ARMOR 



SHIELDS 



The results in bare outline as derived from a statistical treat- 

 ment of the data given in extenso in table 5 are presented in com- 

 parative form in table 6. A perusal of this table brings out the 

 fact that the polyembryonic coefficient of correlation for every 

 one of the five armor shields is in excess of 0.9 and that there is 

 very little difference between the degree of correlation determined 

 for the various regions. One might conclude that correlations 

 of this magnitude are of very general occurrence in connection 

 with polyembryony. Probably the highest coefficient of cor- 

 relation ever determined is that which obtains for the 20 sets 

 just dealt with in the caudal shield, namely, 0.9898. This is 

 remarkably close to absolute correlation and will probably hold 

 the record for such coefficients indefinitely. Further comment on 

 the table is unnecessary as the figures speak for themselves. 



2. UNIPARENTAL CORRELATIONS FOR THE FIVE MAIN ARMOR 



SHIELDS 



In table 7 are given all of the figures used in calculating the 

 coefficients of correlation that obtain between mothers and off- 

 spring for the five armor shields, except those already presented 

 in table 6. It will be noted that, with the exception of the caudal 

 shield, the coefficients average* approximately 0.5. In the case 

 of the caudal shield, however, there is scarcely any appreciable 

 correlation between mothers and offspring. We are therefore 

 justified in assuming that, in this particular lot of fetuses, the 



