200 



The Mechanism of Evolution in Leptinotaesa 



from one small plant, in a very moist situation, and were noted at the time of 

 collection and on measuring the lot these individuals stood out from the rest as a 

 group by themselves. Two months later, in the same location, I collected a sec- 

 ond lot of 1000 (the next generation) and examined the left c spots statistically. 

 The results of this examination are presented in figure 21, where the isolated 

 group found in the first curve of distribution is also present, but in this case it is 

 included in the curve of continuous variations. Another set of observations 

 made at a point near San Marcos, in Vera Cruz, Mexico, gave different data, 

 but results that were identical in principle in that the data of the variability of 



90 

 80 

 70 

 60 

 50 

 40 

 30 

 20 

 10 

 100 

 90 

 80 

 70 

 60 

 50 

 40 



S S 3 



■n 



Fig. 21. — Measurements of left spot c in 

 L. undccimlineata at Tierra Blanca to show 

 differences in the array in two successive 

 generations. 



the left c spot in 500 individuals gave a polygon of distribution with a small 

 group on one side, but connected with it by intergrades. A collection in exactly 

 the same locality in the next generation gave a different condition in the polygon 

 of distribution, with the group standing below the general population entirely cut 

 off, and intergrades were wanting. This extreme group came entirely from an 

 area of coarse gravel with little loam, and dry at all times. The first collection 

 was made in the early part of the rainy season, when it had been rather dry over 

 the plain as a whole, so that there was a rather continuous series of stages of dry- 

 ness from the extreme state of the gravel bed to that of the rich soil of the 



