ONTOGENY OF ADULT COLOR PATTERNS. 159 



pupa. These rules for the order of the development of insect coloration seem 

 to be of almost universal application. This universal similarity in develop- 

 ment, however, is exactly what we should expect to find in animals so defi- 

 nitely metameric as are the insects. 



Of great interest and worthy of a most careful embryological investigation 

 are the centers of coloration which are found in the larvae, upon the various 

 elements which comprise the segment ring. These centers as they appear in 

 the late embryo are first recognizable as groups of cells which are associated 

 with setigerous cells. These are observed to appear first upon the most 

 anterior portions of the body and then to develop progressively caudalward, 

 first upon the sternse, then upon the pleurae, and last upon the tergae. Each 

 center consists of setigerous and specialized hypodermal cells, which before 

 each ecdysis may develop color-forming enzymes. These cells may be differ- 

 entiated from the ordinary hypodermal cells by proper cytological methods. 

 In the earlier stages in the embryo they are very few in number, but they mul- 

 tiply, and at hatching are often scattered thickly over the surface, as in the 

 epicranium and pronotum of most young larva?. In the larval stages and in 

 the ontogeny of coloration at each ecdysis we have seen that the color pattern 

 develops either around these areas or not at all. Of peculiar significance is 

 the discontinuity in the larval patterns; but this is more apparent than real, 

 because after a color pattern has been developed as, for example, that of the 

 first instar there is no possibility of altering it excepting by ecdysis. The 

 cells which are responsible for color formation show, however, no such dis- 

 continuity. Thus, in undecimlineata, the color-enzyme forming cells are 

 active in the late embryo, the full phyletic number of color centers being then 

 developed ; but in the early part of the first instar these cells become inactive 

 in all color centers excepting a few that is, they form no zymogen gran- 

 ules and, consequently, color is not developed anywhere excepting in the 

 spiracula, wing, and baso-thoracic pleural centers, and upon the pronotum 

 and epicranium. All of the tergal, sternal, and baso-pleural abdominal cen- 

 ters remain quiescent throughout the larval life, and become active again only 

 toward the close of pupation. During this interval, however, they increase 

 greatly in number, and late in pupal life are quite generally distributed over 

 the body segments ; so that the imaginal color of these parts is deep metallic 

 black. In other species, as sigiiaticollis, these cells lie dormant during the 

 second instar, but become active in the third, and remain so throughout the 

 pupal stage and in the imago. 



The division of the centers upon the dorsal surface into anterior and pos- 

 terior series is, I believe, due to the development of a fold in the integument 

 of most larvae, which results in the mechanical separation of each center into 

 anterior and posterior portions. In the late embryo the cells are closely 

 bunched, and it is only in the development which follows that the two sub- 



