I'IKUIXAK: KUREMA LISA. 1093 



out stirrinj^ tVom tlie si)ot. Otiiers would uot toutli oitlier. I conclude, 

 therefore, that Cassia alone is the food plant of the larva, and that it has a 

 decided prct'crcucc for the species with small and finely divided leaflets. 



Habits of the caterpillar. After escaping from the egg, the little 

 caterpillar generally continues to eat some of the shell it is leaving, and 

 sometimes devours nearly the whole of it. From the very first it feeds on 

 the under side of the leaves, and eats long, parallel and narrow holes en- 

 tirely through between the veinlets, after the manner of Eurymus. As 

 the leaflets of the food plant close at night, this position would seem to be 

 a necessary one, and also probably induces a habit of feeding only by day. 



\Yhen not feeding, the caterpillar invariably stretches itself out at full 

 length, either along the stalk of the plant or the middle rib of one of the 

 leaflets, where, being of the same color as they, and the stigmatal stripe 

 resembling in its straightness and stiflPness the midrib of the leaflets them- 

 selves, its detection is very difiicult. When disturbed, it will raise the 

 front portion of the body barely above the surface, and sway it from side 

 to side in a slow but deprecatory manner ; but if roughh' handled, it will 

 drop from the leaf, spinning a thread and hanging thereby. 



In preparing fur cluysalis, the loop for suspension is found in the in- 

 cisure between the second and third abdominal segments ; but in the 

 chrvsalis it alwavs crosses the middle of the first abdominal segment. 



Life history. The life history of this insect is very obscure and puz- 

 zling. I hiive tried to translate the facts known from the south by our 

 northern experience, and I'eversed the process only to find fresh diflScul- 

 ties. The principal difliculty is to understand how it passes the winter. 

 In Florida it flies nearly the year around, from early in February to late in 

 November. Early in the spring, however, the butterflies are so few in num- 

 ber, until the end of April, that one might take them to be individuals which 

 had passed the winter in the winged state ; and the behavior of the insect 

 in the north, and the probably similar habit in Xanthidia nicippe renders 

 this the more probable. In May, at any rate, a new and fresh brood ap- 

 pears and by the end of the month becomes numerous, and thereafter re- 

 mains so. Fresh material appears to arrive toward the end of July (Dr. 

 Chapman bred one from chrysalis, August 2), and again from and after 

 the middle of September (Abbot bred it September 13), and very fresh 

 specimens may still be oI)tained in Georgia to the end of October. This 

 would give three broods a year, counting the spring butterflies as a part of 

 the September-October brood, though certainly some of the specimens I 

 have seen collected in March seemed fresh enough. 



North of Pennsylvania, or on the northern edge of its range, I cannot 

 learn that the buttci-fly has ever been seen earlier in the spring than the 

 middle of June ; June 12 at Mattapan, near Boston, is the earliest, and 

 the specimen was noted as good. The brood which this marks flies only 



