12 DISTRIBUTION AND DISPERSION OP LEPTINOTARSA. 



is true, then continuity and directness of variation are no longer of any great 

 value in the determination of relationships and the phylogeny of animals. 



In Lcptinotarsa the lines of dispersal are short and not any too clearly 

 marked. This is due to the fact that the habitats of the various species are 

 only a few miles apart, and that they have been produced, not by wide sepa- 

 ration, but by unlike topographic and climatic conditions situated in the 

 main very close together. It is not possible, therefore, to say with certainty 

 that lines of dispersal exist, except in one or two cases. The lines of dis- 

 persal which can be determined, those of intermedia and decemlineata, point 

 southward to southern Mexico as a center of origin. 



The seventh criterion, "location of least dependence upon a restricted 

 habitat," does not appear in this case, at least, to be of any great weight. 

 For example, L. decemlineata in the eastern United States is less dependent 

 upon a restricted habitat than is any other member of the genus. That this 

 is true is shown by the fact that it regularly feeds upon some dozen or fifteen 

 species of food plants and has been found upon about forty. No other mem- 

 ber of the genus has so many food plants, nor does decemlineata in its origi- 

 nal habitat in Colorado, western Kansas, and Nebraska. Likewise it occurs 

 on dry or wet or sandy soils, on clay or rich alluvium, on plain, hillside, or 

 mountain slope, in the open, and even in the forest. In the case of this 

 species least dependence upon a restricted habitat is a new character (habit) 

 and not an original one. In its original home it feeds almost exclusively upon 

 ,S. rostrahim and lives in the open country (savanna and steppe formations) ; 

 never upon mountain slopes nor in wooded areas. 



Other members of the genus known to be nearer to the point of adaptive 

 radiation, as nndecimlineata, have only one species of food plant and are 

 restricted to a very narrow habitat the open portion of the country. The 

 genus is never found upon mountain sides or in woodland formations. 

 Likewise mnltitceniata, the direct ancestor of decemlineata, feeds only upon 

 5. rostratum and its allies, and does not live upon any other. In this case 

 ' ' least dependence upon a restricted habitat " is a late acquisition of the 

 species and not an original character. 



That Adams's seventh criterion is often one of value is certain, but the 

 case cited above, and others equally good that can be brought forward, would 

 serve to weaken seriously one's confidence in this criterion as one of general 

 value. It might be useful in specific cases, when the other evidence and 

 the history of the group are so fully known that the recent acquirement of 

 this character is beyond question, but when we know this much about the 

 animals in question, the criterion is no longer needed in the search for the 

 center of adaptive radiation. It would be already known from more reliable 

 evidence. 



The eighth criterion, "continuity and directness of individual variation 

 or modifications radiating from the center along the highways of dispersal," 



