150 COI.ORATION IN I.EPTINOTARSA. 



In all of the species in which the ontogeny of larval coloration has been 

 described, and in all in which it has been studied, there appear sharply marked 

 stages, some generalized, others highly specialized. We find, also, that cer- 

 tain color centers are exceedingly persistent, and others transient; that, in 

 general, the tergal and sternal centers are the less persistent, and that the pos- 

 terior members are more variable in their degree of permanency than the ante- 

 rior. The sharp demarkation of stages is, of course, due to the phenomenon 

 of ecdysis, which periodically removes one entire set of coloration characters 

 and affords an opportunity for them to be replaced by a similar set or a differ- 

 ent one, as the case may be. In the phenomenon of ecdysis there are at work 

 processes of high importance. Thus, in the first instar in L. diversa, there 

 are groups of metamerically arranged cells which develop the necessary 

 enzyme for the production of the color centers, and, accompanying them, are 

 cells which develop spines; but at the first ecdysis these spines are all 

 removed, which means that the setigerous cells hold in abeyance their power 

 of forming spines. Likewise, the cells which develop the enzyme necessary for 

 the formation of the many centers of coloration in the first stage do not exer- 

 cise this function in the reconstruction of the color pattern during the first 

 ecdysis, and only the spiracula, wing, and basal-pleural thoracic centers retain 

 this function through the first ecdysis. At the second ecdysis, when the sec- 

 ond color pattern is removed and the third is developed, color-enzyme forming 

 cells, inactive in the formation of the second pattern, become active and pro- 

 duce the dark tergal markings of the last larval stage. These tergal markings 

 develop first about the primary centers, but they are soon joined by deposits of 

 pigment, and thus form the anterior and posterior bands. 



The pigment and its formation is the same in both sets of colors ; but why 

 does color appear in some centers in the first stage and disappear entirely in 

 the second, and then reappear in the third stage in increased area ? Here are 

 fundamental processes which will be considered at some length later on. 



As has been stated, all traces of the cuticula colors are removed at ecdysis, 

 and only the general body color is left to the larva. On its emergence from 

 the old larval skin the larva is yellow, red, or some other color, depending 

 upon the color of the body, and is without a trace of dark markings, excepting 

 the pigment of the ocelli on the head. The first dark colors to develop are the 

 spiracula and wing spots, which are a homodynamous series of color centers, 

 and the color in these areas may even begin to form before the larval skin is 

 cast off. The color changes have been described elsewhere and need not be 

 restated. Color next appears in the appendages of the head — first in the 

 mandibles, next in the palpae and other mouth parts, then in the legs, and 

 finally in the baso-pleural centers. In figs, i to 4, on plate 18, is figured 

 the development of the second color pattern of signaticollis, and in figs. 5 to 8, 

 on the same plate, that of the third color pattern for the same species. 



