Analysis of Heterogeneity in the Population 291 



coco is next in the series, and the Tlalnepantla location least diverse in its 

 population. The diversity is indicated in another way by the occurrence of 

 biotype 1, which may with certainty be expected in the population of the 

 Chapultepec colony. At Texcoco it is about a 1 to 1 chance that one will find it 

 in the population, while at Tlalnepantla, the chances are far less, on the average 

 10 to 1, against finding it present. With reference to the other biotypes much 

 the same set of chances might be derived, differing of course in the chances and 

 in the locations. 



That these three locations, situated on the rim of this great basin, presented 

 constant differences in the character of the population was at first surprising, 

 and when was added to this the weight of the determination from other locations 

 in the basin, giving still other differences in the character of the population, so 

 that at no two points in the whole area was I able to find like conditions in the 

 population of this species, it became increasingly clear that the phenomena were 

 well worth analysis. There has been little trace of migration of the population 

 from one colony to another, although this has been sought for most diligently in 

 the years of the study of the locations. Especially in the close of the season, in 

 the interim between the maturing of the second summer generation and its 

 entrance into hibernation for the dry season, there was opportunity for dis- 

 persion of the population at any colony to nearby locations, producing thereby 

 a mixing and keeping the population of the valley fairly uniform. Any move- 

 ments would, on account of the topography, be almost entirely confined to the 

 valley of Mexico, and the only probable chance of foreign individuals would be 

 from the plains of Apam entering through the pass at the northeast comer of 

 the valley. All other sides were well cut off by topographic or other barriers 

 that in the main prevented either entrance or exit to this basin. One exception 

 to this must be noted in that, in especially favorable seasons, when the country 

 to the north of the basin has had high temperatures and plenty of rain, there 

 might be produced opportunity for food and dissemination, such that there 

 might be entrance or exit on the north. However, I have not been able in two 

 seasons of good conditions in these northern districts to discover any movement 

 in either directions, so that I have come to regard the basin as one that is prac- 

 tically physiographically isolated as far as this form is concerned, as it is 

 known to be for other organisms, especially those of aquatic habitat (Meek). 

 It seems from all that I have been able to find out about the dissemination of 

 these forms in the valley with respect to the regions about, that there is relatively 

 little interchange as a regular thing, and only in exceptional years is there pas- 

 sage of many of the barriers to dissemination. 



Within the valley there is of course more or less in the way of interchange of 

 materials between the nearby locations where the animals are able to grow, 

 although this is not as extensive as might at first seem probable. In the north- 

 ern species, L. decemlineata, it has long been known that the autumn is a time 

 of extensive dissemination by means of flying, and this is influenced by the 

 winds to a considerable degree as has been shown in an earlier publication. In 

 this species, in which the flying habit in the end of the growing season is also 

 developed, I was interested to find if it there too played any considerable part in 

 the dissemination of the animals from one location into another. Observation 

 in the field established the fact that the forms did fly more readily at this season 

 than at any other, and in captivity they do the same, so that the behavior is not a 



