Febbuaey 24, 1911] 



SCIENCE 



309 



individuals to be used in breeding should have 

 at least one square meal a day, served either in 

 the early morning- or evening, when they can 

 be handled without danger of escape. In case 

 they are actively flying during the day, and 

 visit the moistened bouquet provided for them, 

 they will of course feed themselves, but they 

 are more strongly attracted toward the source 

 of light than toward food, so that, if only 

 slightly active, they may not reach the food 

 supply at all. This is one reason against the 

 use of large cages. They do not go in search 

 of food, as the bee seems to do, but, stimulated 

 by light to activity, they find it almost by 

 accident. Hence they must be kept near 

 the food. 



The same reason applies to the use of small 

 cages for mating. The larger the cage, the 

 smaller are the chances that two individuals 

 will meet. I began my experiments by turning 

 butterflies loose in a large screened veranda, 

 strongly lighted only on one side. Under 

 such conditions their attraction towards the 

 light absolutely controlled them. Each went 

 his or her own way, paying not the slightest 

 attention to the others. If several males and 

 females of the same species are placed in a 

 cage of the dimensions noted above (10 inches 

 high and 15 inches square) or 15 inches in all 

 three dimensions, and kept in the direct sun- 

 light, or, if the temperature is high enough, in 

 strong diffuse light, some matings may be ex- 

 pected. As soon as a couple are mated they 

 should be removed to a separate cage, and 

 their numbers noted. Mating continues in 

 Golias philodice for over an hour, usually for 

 about an hour and a half, and often two or 

 three hours, so that there is little danger of 

 promiscuity when large numbers of both sexes 

 are placed in the same cage, if properly 

 watched. One male often can be mated on 

 successive days with several females. 



Fertilized females often begin to lay on the 

 day after mating. They will lay quite well in- 

 doors under warm, sunny conditions, provided 

 they are well-fed, but in any case the food 

 plant on which they lay should be growing, so 

 that the minute larvse, upon hatching, will 

 find fresh food awaiting them. The most con- 



venient way to obtain the eggs and rear the 

 larvse of Golias philodice, for example, is to 

 place the butterfly in a cage over healthy 

 clover growing on a lawn that is free from 

 ants and slugs. 



The food plants and feeding habits of larvse 

 are so varied that few general directions can 

 be given for their care. Caterpillars that feed 

 on coarse leaves that wilt slowly when gathered 

 may be kept in large, flat, cloth boxes, like 

 those that are used at the Gypsy and Brown- 

 tail Moth Laboratory at Melrose Highlands, 

 Mass. But the leaves of most food plants of 

 butterflies soon become wilted and dry under 

 such conditions, and it is preferable to enclose 

 the growing plant in a cage, or a branch of the 

 shrub or tree in a bag of cheese cloth. If kept 

 in close glass vessels to prevent evaporation, 

 or even in large moist vivaria, there is danger 

 in Colias philodice, at least, of infection with 

 an intestinal bacterial disease that may kill a 

 large proportion of both larvse and pupas. So 

 a good general rule is to keep the larvag upon 

 their food plant out-of-doors, and well 

 screened from insect parasites and birds. 



Pupse should be carefully guarded from mice 

 and slugs, and if either are at all abundant the 

 chrysalids should be taken indoors. Limax 

 maximus will gnaw through a fine cloth net- 

 ting on the inside of which a pupa rests, and 

 eat the chrysalis, leaving a small hole through 

 the screen, the margin of which is blackened 

 with the salivary juices of the slug, which ap- 

 parently contain sulphuric acid. 



The student of the genetics of butterflies 

 will hardly be content with the ordinary 

 museum methods of mounting specimens upon 

 pins stuck into cork. Greater compactness, 

 perfect safety from museum pests, and quick 

 accessibility to both surfaces of the wings are 

 needed. The patent Denton mounts for indi- 

 vidual specimens satisfy these requirements 

 well, but they are expensive for use in large 

 numbers and, though vastly superior to the 

 ordinary pin-cushion method, are not conve- 

 nient to handle when huge families are being 

 examined and compared, because each speci- 

 men must be picked up separately. All these 

 needs, however, are met by a simple form of 



