BUTTEKFLIES OF THE DISTEICT OF COLUMBIA 49 



are present or not are unable to take advantage of the puddles and 

 muddy spots and become more or less evenly distributed over barren 

 areas devoid of the food plant and of females, being in this area 

 most conspicuous along the roads. 



BUTTERFLIES AND STORMS 



In the drier portions of the Tropics, heavy showers following a 

 period of dry weather are succeeded in a few hours by the appear- 

 ance of great numbers of butterflies of various kinds, and in cer- 

 tain places this phenomenon is very striking. 



In summer in the vicinity of Washington the best collecting is 

 always to be found on the day following a severe thunderstorm, and 

 on the day succeeding. On these days the insects are mostly fresh 

 from the pupa, and good series of first-class specimens may be 

 obtained. 



One reason for the excellent condition of the butterflies on the 

 days immediately following a violent storm is that the old, worn, 

 and weak individuals are killed off, only the more vigorous surviv- 

 ing. But this does not seem wholly to account for the facts. The 

 storm itself seems to stimulate the emergence of fresh individuals. 



The emergence of manj'^ of the Lej^idoptera can be greatly delayed 

 by keeping the pupae in dry surroundings, restoration of the normal 

 conditions causing them to appear promptly. In the case of the 

 Ailanthus silkworm moth {Philosainia cynthia)^ I have obtained 

 perfect adults from cocoons that had remained dormant, under con- 

 ditions of abnormal dryness, for more than 20 months. 



The effect of a heavy rain is undoubtedly the result of the great 

 increase in the supply of available water, the additional moisture 

 enabling the butterflies immediately to complete their transforma- 

 tion to the adult — or rather to escape from the pupa. 



Though the increased proportion of fresh butterflies after a rain 

 in summer is often very evident and obvious, the actual number of 

 individuals on the wing is as a rule not appreciably altered, so that 

 the effect of a storm is not apparent to a casual observer, and indeed 

 is not satisfactorily demonstrable. 



The hot and exceptionally dry summer of 1930 furnished an ex- 

 cellent opportunity for making observations on this point. 



On September 15 the meadows beyond Cabin John, Md., were 

 thoroughly explored, and the various species present were noted. 

 All butterflies were unusually scarce. The milkweed butterfly {Dan- 

 aus plexippus) especially was infrequent — I saw only four or five. 



On the afternoon of the following day (September 16) there was 

 a heavy shower, the first for many days. 



