652 JOURNAL, BOMBAY NATURAL HIST. SOCIETl, Vol. XXV. 



in tlie wild state, they are laid singly .... on account of the 

 cannibal habits of the lai'va^ ". In the process of rearing the larvfB 

 he further remarks : " All the eggs hatched on July oOth, remain- 

 ing six days in the egg state, the larva makes its exit by eating 

 the crown as well as a portion of the side of the egg-shell. Direct- 

 ly after emergence it is exceedingly small. It is very active 

 when first hatched, crawling rapidly for such a small creature and 

 feeds on various parts of the plant (Lotus coruiculatus). I found one 

 feeding with its anterior half buried in the end of a stem wdiich had 

 been cut off. First moult August ord, the first stage only lasting 

 four days. Shortly before the first moult it measures only 1*6 mm. 

 long. During the early stages these larva?, like other Lifca'nida/, 

 require the greatest attention to distinguish their changes, on ac- 

 count of their small size and similarity of stages, and unless most 

 carefully watched under a lens it is practically impossible to 

 detect their moults. Second moult, August 7th, In this stage it 

 greedily feeds on the seeds of Medicago lupuUna, eating through the 

 capsule, and devouring the contents, but appear to feed mostly at 

 nights. Third moult, August 12th, In this stage they feed 

 as much by night as by day but are much slower between 

 the moulting, occupjdng eleven da^^s from the third to the fourth 

 moults. Fourth and last moult, August 23rd. Although the gland 

 is so similar to the gland oi Lijcana arion, I have been unable to 

 detect any liquid exuding from it when touched ; also it appears 

 less sensitive when touched. During the last stage I supplied the 

 larvffi with both the common white and red clover-blossoms, 

 which they greedily devoured. One larva completely ate up the 

 whole of the petals of a red blossom in two daj^s ; it appeared to 

 be feeding continually both day and night the whole time. It 

 also feeds on the flowers, seeds and leaves of Lotus cornievlatus. 

 During their earlier stages the larva3 are decidedl}' cannibalistic, 

 I noticed one after the first moult feeding on a newly-hatched larva, 

 which it seized as it emerged from the egg ; but during the last 

 two stages I did not find them attacking each other, which agrees 

 with the cannibalism of L. anon larvae. The pupa is attached 

 to the foodplant by the cremastral hooks to a pad of silk and a 

 girdle round the bod}^ ; all nine were attached to the centre of 

 the leaves, and the general resemblance, both in form and colour, 

 between a decaying Jjottis leaf and the piipa is very great. The 

 pupa remains in that state from about ten to fourteen days accord- 

 ing to temperature. The first imogo emerged on September Gth, 

 the last on September 18th, 1904," 



Scudder saj'-s that, in Europe, the larva feeds on ' ^ Lotus, Avthyl- 

 lis, Medica'jo, Trifolmm, Pisum and Onohrychis, and even on Fham- 

 nus; in North America it feeds ou Les2>edeza,Flias€olns, Desmodiiim, 

 (taladia, Trifolium and Astrayalus.^' 



