146 COI.ORATION IN I^^PTINOTARSA. 



segments, however, the elementary color pattern is retained in full, as may be 

 seen by comparing fig. i on plate 17 with text-figures 7 and 8. Each color 

 center bears one or more single or slightly branched spines on the tergal and 

 sternal sclerites of each segment. In the pleural region these spines are 

 small, and are limited to the baso-pleural centers. As the small larvae push 

 about through the abundant trichomes on the leaves of their food plants, a 

 large accumulation of these become lodged among and cemented to the 

 spines by the secretion of the dermal glands, until the larva presents the color 

 and appearance of a ball of dislodged trichomes. At the first ecdysis the 

 color pattern of the larva changes strikingly, as is shown in fig. 2 on plate 

 17. In the second stage the epicranium and pronotum remain black, while 

 all the color centers are lost, excepting the abdominal and prothoracic spirac- 

 ular spots, the wing spots, and the baso-pleural thoracic spots. The integu- 

 ment in this second stage is smooth and entirely devoid of spines on the tergal 

 and sternal elements of the segments. Owing to the sticky secretion of the 

 dermal glands a deposit of trichomes gathers on the larva in this stage, 

 though, owing to the absence of spines, this deposit is thinner than in the pre- 

 vious stage. The true body color now is a pale, transparent greenish- white, as 

 is shown in the figure. Between this stage and the final one shown in fig. 3 

 on plate 17, there are no modifications in the color markings, although the 

 general body color is changed to an opaque, pale yellowish-white, which is 

 due solely to the development of the fat body. In the final stage the increase 

 in the body surface, which is not accompanied by any great increase in the 

 number of dermal glands, results in there being on the body only a very 

 slight deposit of trichomes, if any at all, so that the larvae are freely exposed 

 on the leaves of their food plant. The final body color varies from a pure 

 white to pale lemon yellow, the average condition being represented by fig. 3 

 on plate 17. 



The nearest relative of L. undecimlineata is L. dwersa^ and it is of interest 

 to compare the ontogeny of the larvse of the two species. In the first instar 

 diversa has the appearance and all the characteristics of a larva of undecim- 

 lineata at the same stage of development. In fact, the larvse of the two spe- 

 cies are then alike, as may be seen by comparing figs, i and 4 on plate 17. 

 During the first instar L. diversa has likewise the habit of covering itself with 

 the trichomes of its food plant, so that the two can not be distinguished from 

 each other either in coloration or in habits. The second instar is, as in tmde- 

 cimlineata, very unlike the first, there being a striking loss of color centers 

 and of spines, while in color and habits the two species are at this stage almost 

 duplicates of each other, as may be observed in figs. 2 and 5. In the final 

 stage, however, as is shown in figs. 3 and 6, there appear striking differences 

 between the two species in color and color pattern. L. diversa now develops 

 a series of black tergal markings that are characteristic and constant, and 

 which clearly differentiate the two species of larvae. The black tergal mark- 



