REVISION MALACOSOMA HUBNER IN NORTH AMERICA 65 



possible to rear any of the 1 96 1 Fj egg masses because of the difficulty 

 in holding them over the winter. The eggs, however, went through 

 their embryological development almost as well as the 1960 F, egg 

 masses despite the fact that conditions may have been less favorable in 

 1961, and there is no reason to believe they could not have been as 

 successfully reared. This is true of all populations studied in 1961, 

 including those which are regarded as M. incurvum. Incurvum, however, 

 is regarded as a good species for other reasons which are discussed in a 

 later section (see page 118). 



The areas showing great larval variability in the field were usually 

 quite extensive and, in general, they must be regarded as zones of 

 intergradation between populations of the same species. Some popula- 

 tions are sufficiently different in one or more life stages, and well 

 enough defined geographically to warrant recognition as subspecies, 

 but recognition of them as good species would completely obscure the 

 relationships between them. 



It seems possible that these populations, which are presently quite 

 different in some respects (most often in larval color pattern) yet almost 

 identical in many other respects, could have evolved these 

 differences in the past when they were separated physically by a barrier 

 of some kind. If this is true, the highly variable and extensive popula- 

 tions which presently occur between them must have been formed more 

 recently when they came in contact as the barriers disappeared or were 

 modified, and before reproductive isolation had become well developed. 

 Any suggestions as to where or when these populations were separated 

 and reunited would largely be pure speculation, so they have not been 

 attempted. 



Another possible explanation for the present differences between 

 some populations and the relatively large zones of intergradation 

 between most of them is that they are the visible expression of gene 

 combinations that are most favorable under the prevailing conditions 

 in each area, and, therefore, are preserved by natural selection. This 

 would not require that they be physically separated in the past. 



The most likely explanation, however, for the differences is one 

 that involves at least partial physical separation at some time as well as 

 the preservation of favorable gene combinations by natural selection, 

 since both of them acting together would be more effective than either 

 one alone. 



