ii6 



We find, therefore, a correlation between the thoracic segments and 

 the three last abdominal segments, which is especially marked in 

 Papilio Phileno}\ where the warts are prolonged into horn-like ap- 

 pendage's. 



A further instance of this correlation is afforded by Danais Archip- 

 piis and Lmietiiiis Disippus, also from West Virginia, both of which 

 I have carefully examined. On the first-named species two long horns, 

 quite similar to those on Papilio Philenor, stand on the second thoracic 

 ring, and very similar ones on the next to the last abdominal ring; in 

 fact, rudiments of these are conspicuous even in the first stage. Li- 

 mcnitis Disippns has horn-like warts on several of the segments, the 

 most prominent being found again on the second thoracic and the last 

 abdominal ring. They are not, however, equally large, as in Danais 

 Archippus, but the horns on the second thoracic ring are by far the 

 longest, and these are undoubtedly the ones which have a purpose to 

 subserve, perhaps as a means of inspiring terror. The warts on the 

 other rings only seem to repeat the organs referred to, just as we some- 

 times see in segmented animals characters of one segment carried over 

 on the others. In Limenitis Arthemis, according to Edwards, the 

 horns on the second thoracic ring appear to be decidedly larger in 

 proportion to the other warts. 



To return to the Papilionidce, the normal course of the warts is to 

 decrease in size after each successive moult, and sooner or later to 

 vanish entirely. Those standing on the middle segments, as they were 

 the smallest from the start, are the first to disappear, while those on 

 the thoracic rings and the last abdominal rings remain the longest. 



Hence we conclude that the warts with their bristles lose their mean- 

 ing for the caterpillar, and are therefore abolished. If we observe the 

 stages of . development of the Papilio caterpillars on plate VII, we 

 shall at once notice that, with the exception, of the easily distinguished 

 Papilio Philenor, these warts, even as early as the second or third 

 stage, are to be set down as rudimentary organs. Their original inten- 

 tion is to form suitable and prominent points of support for the bristles, 

 which at all events must have a meaning for the caterpillar. Whether 

 they have any meaning in the Papilionida, however, seems to me 

 questionable; at any rate this is the case only in the earliest stages. 

 On the whole we are inclined to believe that the warts and the bristles 

 attached to them may be an inheritance from ancestors for which these 

 features had an important purpose, as is undoubtedly the case with 

 the Nymphalidce. The warts become rudimentary in proportion as 

 the markings come out on the body of the caterpillar; in other words, 

 the markings take the place of the warts, as both together would evi- 

 dently have the effect to interfere with each other. Natural descent 

 chooses as the objective point of its creative skill the markings, and in 

 many cases the bright coloring of the larvae of Papilionidce. Other 



