STUDIES OF VARIATION IN INSECTS 207 



Fitch's estimate of twelve million cherry aphids to the tree in a 

 badly infested orchard, and Doten's ground-covering armies of 

 western crickets in eastern Nevada, with two miles of rank front 

 and a thousand yards of file depth. Yet with all this numerical 

 wealth of kinds and individuals, insects show an amazing stead- 

 fastness of fundamental structural plan and a striking consist- 

 ency of physiological processes. This brings it about that in- 

 sects compete sharply with one another in the struggle for place 

 and food. At the same time the abundance of individuals insures 

 a wealth of small variations, and an increased chance for larger 

 variations, /. ^., *' sports." Thus Natural Selection ought to 

 find a congenial field in the insect class, its two prerequisites, 

 (rt) numerical and heterogenic wealth of variations, and (d) 

 sharp competition, coexisting and strongly emphasized. On 

 the practical grounds, too, of ease of collecting, preserving, and 

 studying large series of individuals, of accurate determination 

 of geographic limits, and of the possession by the subjects of 

 distinctive substantive characters, such as strongly colored, 

 sharply delineated pattern, as well as readily counted or meas- 

 ured meristic characters, as tarsal and antennal segments, 

 spines, hooks, hairs, etc., the insects are peculiarly advantage- 

 ous subjects for variation studies. Finally, many insect species 

 with quickly succeeding generations may be readily bred, gen- 

 eration after generation, in the vivarium or laboratory, under 

 practically natural conditions, or under conditions of controlled 

 varying nutrition, temperature, light, humidity, etc., thus afford- 

 ing opportunity for studies in heredity and for experimental 

 studies demanding accurately determined controlled changes in 

 conditions of life. 



But most important recommendation of all is the fact that the 

 unique character of the post-embryonic development obtaining 

 among the more specialized orders of insects affords an oppor- 

 tunity to distinguish sharply between variations which must be 

 held to be strictly blastogenic (congenital) and others which 

 may be in large part acquired. The immense desirability of 

 being able to distinguish between strictly congenital and ac- 

 quired variations is of course obvious to all students of variation, 

 heredity and species-forming. If acquired characters (acquired 



