Analysis of Heterogeneity in Some Simplest Chaeacters 231 



Three lines shown in table 33 showed distinct tendencies to become restricted 

 in their ranges, and these three, 10 g VIII (BAAA), 10 g VIII (BACA), and 

 10 g VIII (BADA), and were preserved for further analysis to see if continued 

 operations of the same kind would further restrict the range in the fluctuations 

 of the index. The remainder of the lines were killed off or used for other work, 

 having given no indications of curtailment of the range as a result of the opera- 

 tions. The three lines reproduced between the middle of March and the middle 

 of June, gave two generations in which the same standard of parental indices 

 were maintained in both generations, with the result that at the end of the tenth 

 generation in June 1908, three clearly defined groups had resulted, and which 



Table 33, 



No, of mating. 



Parents, 

 index. 



Progeny. 



Fraternity No. 



Range of index. 



Totals. 



8. glV(AA) .... 

 10. glV(BA) .... 



8. gV(AA) 



10. gV(BA) .... 

 M. 2. gIV 



M. 2. g V 



2.625 



2.770 



2.753 



2.701 



2.650 

 2.761 



2.883 



'2.681 

 2.891 

 ^2.714 



2.851 



2.741 



2.761 



2.626 



2.741 

 2.789 



2.675 



2.741"! 



2.814 y 



2.762J 



8. g V(AA) 

 10. g V(BA) 

 10. g VI(BA) 

 10. g VI(BA) 

 M. 2. gV... 



M. 3. gVI.. 



(M. 

 IF. 



{f. 



72 



75 



103 



133 



Total 516 



could no doubt have been maintained in the same condition and range of index 

 as long as required. The differences between the three strains were so small that 

 it was not possible to certainly distinguish them in the cultures by inspection, 

 and only measurement could reveal the true state of affairs. 



The condition in these three strains continued through the four generations 

 that they were carried, showing that there were minor differences in the original 

 " biotypic group " which could be isolated by more minute methods of analysis 

 and perpetuated by common methods of isolation, proves that the original 

 " fluctuations " are not the simple set of " environmental " disturbances they 

 are sometimes asserted to be. 



In the autumn of 1908, after hibernation since June, the three cultures were 

 again put to breeding, and two more generations of precisely the same character 

 were produced, indicating further the constant nature of the small differences 

 that were being dealt with. In the last three generations it was found that 

 there was a tendency for the groups with the higher and lower indices to 

 transgress the bounds of the variability of the original " biotype " stock, espe- 



