443] COLORS OF TIGER BEETLES— SHELFORD 49 



of moisture from the surface film and are in the opposite direction as 

 compared with the ontogeny changes and the changes which take place 

 on drying of fresh immature specimens. The cause of these physical 

 changes is unknown. Other shiftings in color have been noted; one 

 of these in C. lepida is of particular interest as the change is in the 

 direction opposite to that already noted. C. lepida. has the elytron 

 nearly all white but such parts as are pigmented are green in the 

 adult. When the pigment begins to develop it is a brilliant gold and 

 remains so for several days, finally changing to a dark green. In this 

 case the change is the only one of the kind noted. Golden yellow 

 blending with green is commonest in cuprascens sperata, circumpicta, 

 and related species. These may shift from green to brown through 

 yellow instead of red but their ontogeny has not been studied. 



Even the dull species like C. 12 ^guttata, and repanda, and occa- 

 sionally C. punciulaia show more green in the early stages and turn 

 brown as they mature. The early stages of C. tranquebarica are black- 

 ish green, graduallj- turning bronze brown as more pigment is devel- 

 oped. C. formosa is at first reddish and gradually changes to brown ; 

 some individuals collected in the autumn are red. 



RELATION OP ONTOGENETIC STAGES TO GEOGRAPHIC RACES 



First of all it should be noted that there appears to be no good 

 reason for assuming that the biogentic law holds good with reference 

 to these color changes; it would be only the most radical adherent who 

 could see it applying. However in a general way the developmental 

 stages of a given species, like C. scutellaris lecontei, may practically 

 reproduce the color of another variety in ontogeny. Compare figure 

 546 with figure 553, plate XXIX. The stages in the development of 

 typical C. purpurea do quite exactly duplicate some of the races rec- 

 ognized. Tlius the stage shown in Figure 563 practically duplicates 

 the color of graminea while Figure 564 duplicates some of the speci- 

 mens of the subspecies 10 notatta.^ 



Again in Figure 572 appears an ontogeny stage which resembles 

 very closely the variety denverensis but is less yellowish; denverensis 

 also usually lacks the blue green margins, though the green is purer 

 and brighter along the margin, showing a difference comparable to 

 that seen in nearly aU specimens of purpurea. 



C. hiriicollis shows a stage in the development of color (Figs. 

 559 to 562, PI. XXX) which corresponds very closely to the race of the 

 species occurring on the Pacific Coast. In addition to this it shows 

 slight Inflections of the bluish of the bluish drab forms of Vera Cruz. 

 Reddish brown forms occur in southwestern Kansas. 



Occasional specimens of C. tranejucbarica collected in Massachu- 



