l6o COLORATION IN LfiPTINOTARSA. 



centers become differentiated. This separation of the centers into subcenters 

 upon the tergal elements is, however, a constant feature of insect coloration, so 

 that it must be a fundamental one which, as far as I am able to see, can have 

 no other significance than that it is due to a mechanical division of each seg- 

 ment by a median fold. It appears to me to be entirely explainable on the 

 basis of physiology and ontogenetic adaptation, and to be without any phy- 

 logenetic value. 



All the evidence derived from study of the embryonic development of 

 these centers points to the conclusion that they are true metameric structures 

 which are produced by the successive division of mother centers as successive 

 segments are marked out, and that the mother centers are probably derived 

 from one or a group of blastomeres, yet to be distinguished, which arise early 

 in the embryonic history. These homodynamous groups of cells I have been 

 able to recognize very early in embryonic life, often at the time of the forma- 

 tion of the limb buds, and always soon thereafter. The existence of these 

 homodynamous centers, their relation in later life to sclerite formation, 

 attachment of muscles, and to the later color patterns, show them to be 

 structures a knowledge of whose development and behavior is all important 

 to the proper interpretation of insect coloration. 



These centers or groups of color-enzyme forming cells may in the ontogeny 

 of the insect become quiet, developing no zymogen, they may be entirely lost, 

 or they may become extremely active and increase in numbers until they come 

 to occupy larger or smaller areas in successive stages of the ontogenetic his- 

 tory. It is with these stages and the laws which they follow that we have to 

 do largely in the study of color-pattern ontogeny. Upon the wings which 

 form so conspicuous a portion of insect coloration there are other systems of 

 fundamental coloration which we shall discuss presently. 



During the ontogenetic development of these insects the behavior of the 

 color centers follows closely along definite paths and in rather regular se- 

 quences. Thus, in the larvae, it has already been shown that the spiracula 

 and wing areas are the most permanent and least variable ; that the middle, 

 tergal, and sternal are the least permanent and most variable; and that the 

 posterior members of each series are more variable than the anterior. In the 

 ontogeny of the imaginal color patterns we have seen that the color develops 

 around these centers, and in no other way. Thus, on the pronotum, color 

 appears first at points homologous to the inner and outer tergal centers, and 

 later in the middle tergal. Moreover, the inner tergal centers are fre- 

 quently united to form the spots a and a', while the middle and outer tergal 

 centers are united anterio-posteriorly, but less frequently, there being a more 

 frequent union between the middle and outer series, or between d and e or 

 o and b. A comparison of the figures given shows the equivalence between 

 the color ontogeny of the pronotum and that on the abdominal segments. It 



