DISTRIBUTION OF LINKATA GROUP. 21 



LEPTINOTARSA RUBICUNDA. 



At Toluca, Mexico, at an altitude of 8,500 to 9,000 feet, is found another 

 form, derived from multitaniata, with which it lives. This form {rubicunda} 

 is, however, able to maintain itself only within a narrow area as a member 

 of the fauna. Its distribution is shown on plate 1. As far as I am able to 

 determine, the habitat of this form is limited to the flat plain about Toluca, 

 where it lives upon Solanum rostratum in physiological isolation from mul- 

 titceniala. It is able to reproduce itself without change, as I have proven 

 by breeding it for six consecutive generations. This form is a developing 

 species, as is shown by the fact that my specimens of multitceniata from 

 Toluca have given me as an extreme variation a form exactly like the 

 rubicunda which I obtained at Toluca. These extreme variations were 

 almost completely sterile when crossed with the parent species, but freely 

 fertile when crossed with the rubicunda strain from Toluca. This experi- 

 mental evidence I shall present in detail in another chapter of this paper, 

 but the facts above noted are strong enough evidence to serve as a basis for 

 asserting that rubicunda is a species which is able to maintain itself as an 

 independent member of the fauna of its habitat. Whether it can do this to 

 the extent of widening its boundaries and adapting itself to slightly changed 

 conditions only time, careful observation, and experiment can determine. 1 



Distribution of Leptinotarsa intermedia and decemlineata and Probable 

 Movements within Historic Time. 



leftinotarsa intermedia. 



Over the central table-land in the valleys of the Rio Panuco, the Rio 

 Grande de Santiago y Lerma, and the Rio Grande systems, and throughout 

 southern and western Texas, New Mexico, and Arizona occurs, sparsely 

 distributed, another form, to which I have given the name intermedia. Its 

 food and habits and the conditions of its habitat are the same as those of 

 multitcEniata, excepting that it occupies a region of much less rainfall. Like 

 multitcsniata, it shows a wide range of individual variation corresponding to 

 its horizontally extensive habitat. The distribution is decidedly discontinu- 

 ous, due largely to the conditions of climate, and more remotely to topography. 



In the southern portion of its range this form merges with multitaniata 

 and to the north with decemlineata. Over most of its range it is in life dis- 

 tinguishable from either of these species by constant characters in both larva 

 and adult. 



The distribution of this form within the area occupied by it is not to any 

 great extent controlled directly by topography, because everywhere proper 

 conditions of topography and soil are present. The relative humidity and 



1 The distribution of L. rubicunda was in September, 1906, limited to a tract of about 

 10 acres in extent, where only a very few specimens were found. In August, 1903, it 

 was found over an area of some 7 or 8 square miles. Apparently it is being exterminated. 



