Analysis of Hbtbeogeneity in the Population 371 



The observations were begun in 1904 at this colony and ended at the close of 

 1909, covering 12 generations in the same location, and are lineal descendant 

 populations, with probably some additions of new blood from outside of the 

 colony. The records show the response of the population as a whole at this 

 location to all of the conditions it had to meet, and as registered on the pattern 

 of the pronotum. 



The first census of the population, made in 1904, showed a total count of 3,550 

 in the colony, with the distribution of the diiferent biotypic groups without 

 breaks, and the array when plotted gave complete gradations along all of the 

 lines with extremes lacking, with the exception of the series composed of bio- 

 types 1, 2, 5, and 10. Comparatively little " somatic " variability in the amount 

 of pigment is shown in the population. 



The second census, made in 1904, showed somewhat different conditions, with 

 a population of 1,970 in the colony, and instead of the homogeneous array which 

 the first census presented, the distribution in the second is broken and shows 

 many gaps. In the males there was a clearly isolated group of biotype 1, 

 another of 10 and 13, while the extreme conditions of biotypes 8, 6, and 5 stood 

 apart from the population as distinct groups of considerable strength. The 

 males showed much more variability than did the females where the only dis- 

 tinctly isolated biotype was the extremes of 13. Comparison of the two cen- 

 suses for the year shows quite different conditions for the population, and illus- 

 trates the utter uselessness of a single determination of the condition in the 

 population as a means of discovering the true condition therein. These census 

 records I have shown in figures 54 and 55. 



In the third census, made in the early portion of the season of 1905, the 

 population was smaU in number, 1,897, with a distribution much like that at the 

 beginning of 1904, with no separated groups, and the entire range fairly well 

 represented, with the exception of biotype 13, and the curtailment of the 

 extreme conditions of biotypes 5, 6, and 8. 



The fourth census, made in the latter portion of the season of 1905, showed an 

 increase in numbers, 3,330, and the separation of the population into isolated 

 groups, the probable absence of biotypes 5 and 6, the isolation of the extremes 

 of 8 in the males, and the isolation in both sexes of 10 and 11. On the whole, the 

 population in the fourth generation showed much restriction in its range, and 

 this came in a season and the portion thereof when the precipitation was low 

 and distributed in such a manner that the critical periods in the life of the 

 population were passed in relatively dry conditions with .high rates of evapora- 

 tion. The condition in the population for the two censuses of the year is shown 

 in figures 56 and 57, and exhibit about the same condition as in the year 1904, 

 with the absence of extremes, excepting in the series composed of biotypes 1, 3, 

 3, and 4, and the isolation from the mass of the population of the same groups 

 that were isolated in the second generation of 1904. 



In the year 1906 the first census made (the fifth) showed a new arrangement 

 of the distribution of the population, in that the 9, 11, and 13 series was well 

 developed in both sexes and strong in numbers, whereas the groups 5, 6, and 8 

 were practically wanting in the population. The series 1, 3, 3, and 4, while 

 present, were, in proportion to the whole, not as strong as in previous censuses. 

 The only isolated group was a small number of males of biotype 5. This condi- 



