BUTTERFLIES OF THE DISTRICT OF COLUMBIA 53 



the individual, and furthermore the movement usually involves all 

 the normal and healthy individuals of a species in a given area. 



Butterfly migrations, on the other hand, rarely involve more than 

 a portion, and often a relatively small portion, of the individuals in 

 any region. So far as the individuals are concerned they are in one 

 direction only, and so far as the species is concerned there is seldom 

 any indication of a definite return movement. 



There are other differences between the migrations of birds and 

 the so-called migrations of butterflies, but there are also some curious 

 correspondences. 



In some butterflies migration is a regularly recurrent annual phe- 

 nomenon, in others it takes place more or less frequently but at 

 irregular intervals, while in many it is observed only occasionally. 

 In some butterflies it is more or less characteristic of certain forms 

 within a species but never occurs in an alternative form. It is 

 probable that many butterflies not usually so regarded are in reality 

 migratory, traveling always as independent individuals and possibly 

 at night. 



Of all butterfly migrations those that have attracted the most 

 attention have been those of our common milkweed butterfly {Danaus 

 flexiffus). Observations which, by a fortunate circumstance, we 

 were enabled to make on this butterfly, suggest a possible explanation 

 for its interesting and not infrequently spectacular migratory flights. 



In the middle of September, 1930, a heavy rain succeeding a long 

 period of dry weather caused the emergence of this butterfly in 

 great numbers in the meadows west of Cabin John, and the be- 

 havior of the insects was carefully watched (see p. 122). For 

 about three days they devoted themselves wholly to feeding, 

 spending the night among the flowers in the meadows. During 

 this time they became more and more sociable, showing an increas- 

 ing tendency to gather into loose flocks. They then became more 

 restless, and some were observed to rise high in the air and drift 

 away before the wind westward toward Great Falls. At this time 

 they developed a preference for perching on branches of trees rather 

 than simply resting on the goldenrods and other herbaceous plants 

 in the fields as they had previously done. A few days later — just a 

 week after their first appearance — ^they were seen in Washington, 

 and a visit to the meadows showed that all but a very few of them, 

 which were probably younger than those of the large swarm, had 

 disappeared. 



From these observations the origin of a migrating swarm of these 

 butterflies is easily understood. A number were seen to depart high 

 in the air in the same direction. Had the feeding area been more 

 extensive and the insects therefore more numerous, we undoubtedly 



