222 Experimental Zoology 



forms, or mutations mentioned above. This comparison raises 

 the question whether the two forms of variation, fluctuating 

 and discontinuous, are fundamentally different, or whether they 

 are only extremes of the same process. Let us examine this 

 question more closely. 



Lang gives the following imaginary case that bears on our 

 problem. Suppose there existed a species having three varieties, 

 differing according to size, being 4, 6, and 8 centimeters in 

 length. On crossing they Mendehze and do not give intermedi- 

 ate types. Differences in nourishment, however, may cause the 

 size of any one of the three types to be a httle larger or smaller, 

 2 milhmeters being the extreme in either direction, but this 

 will cause no overlapping. Suppose, however, that the fluctua- 

 tions, due to differences in food, for example, are so great that 

 the largest individual of the 4-centimeter variety is as large or 

 larger than the smallest individual of the 6-centimeter variety. 

 In other words, the individuals transgress the hmits of each 

 variety, i.e. the breadth of the fluctuation of the variety is so 

 great that it overlaps the Hmits of the nearest related variety. 

 Then we should have a population in which a continuous series 

 could be traced from one extreme to the other. Apparently we 

 should be dealing with a continuous variation, but in reality 

 with three pure lines that overlap. Breeding experiments alone 

 could determine that these three types actually exist. Such 

 experiments would be comparatively simple where self-fertiHza- 

 tion is possible; but where cross-fertihzation is necessary, the 

 separation of the population into its three true races becomes 

 more difficult. 



Lang has found in Hehx hortensis somewhat similar conditions 

 to those given in this imaginary case. It has been pointed out 

 that while in some colonies only two sharply separated types 

 appear, in other colonies numerous variations are found whose 

 extremes run into each other. Within the species there exist the 

 possibihties of eighty- nine variations in the banding, and the color 

 may be white, yellow, orange and brown, or ash-gray. There 

 are size varieties also, and the bands themselves may be continu- 



