328 The Mechanism op Evolution in Leptinotaesa 



oval or roughly so in outline and on the average were about 60 feet in diameter. 

 They were denuded of the original vegetation, sown with the seeds of 8. rostra- 

 turn from the valley, and in a few days the location was ready for the test. 



Early in June 1905, 1 introduced into the locations prepared random samples 

 of L. multitcmiata from the colonies at Chapultepee, Texcoco, Tlalnepantla, 

 and Chalcicomula, and all thrived and became self-sustaining with the excep- 

 tion of the introduction from Texcoco, which died off, owing to failure of the 

 food-supply. This was the result of too shallow soil in the spot chosen. In 

 each test the same method was followed as was employed in the natural colonies, 

 that of making a census of the population twice each year. At the time I did 

 not know what the condition in any of the colonies would be shown to be, but I 

 was certain from previous experiences with the biometric methods of study that 

 the problems had to be attacked, and that this was the best method in this one 

 direction, so that in the series the test and the study of the conditions in the 

 colony went on simultaneously, a combination that has obvious advantages. 

 The other locations were utilized for other tests of the same sort that were 

 made on other species. Observations were continued at the three locations, 

 two determinations in each season from 1905 to the end of 1908, eight genera- 

 tions in all. One exception in the census determinations was made, namely, 

 the dropping of the rating of the amount of the pigment present in each of the 

 pattern types shown, only the types being recorded in the test locations, because 

 it was thought that the persistence of the types was the point to be determined 

 and not the amount of " fluctuation." In presenting the records of the testa 

 in this paper the sexes are not separated, although they were in the original 

 determinations. There are no sex limitations or dimorphisms with regard to 

 this pattern, and the mixing of the two into one population statement does 

 not, as far as is known, hide or alter any of the conditions of the test. 



THE CHAPULTEPEC COLONY TEST. 



The character of the introduction of the stock from the Chapujtepec colony 

 is exhibited in figure 147 in the array of the population for the first genera- 

 tion of the year 1905, and were the overwintering members of the previous year's 

 last brood. The chance distribution of the parent test group is shown in the 

 figure in comparison with the progeny of the group at the normal location. At 

 the test location the whole population was allowed to take part in the operations 

 of reproduction throughout the series, so that the test series was treated, in all 

 respects, like an independent isolated colony. The results of the test are 

 shown in figures 147 and 148, and do not reveal change in any essential in the 

 test location. The two locations were not entirely identical, and it was 

 thought that there might be some difference in the new location, due to the 

 change, and some little was found, but it is not at all certain that it was the 

 product of the imlikeness in the conditions, and might as well have been due to 

 altered relations of agents in the gametic complex as to the environment. 

 There was certainly nothing at any point to indicate the environment as the 

 sole active agent in producing the observed departures (figs. 147-151). 



