122 BULLETIN 15 7, U. S. NATIONAL MUSEUM 



crossing a large pond northward early in September, evidently flying 

 with the wind. 



In connection with the migratory and swarming habits of this 

 butterfly the following observations are of interest : 



At Cabin John when it appeared on September 17, 1930, in great 

 numbers (following a rain on the preceding day) it was at first more 

 or less evenly distributed over the fields, feeding everywhere on the 

 goldenrod and also on the few and widely scattered thistles. On the 

 next day (September 18) conditions were about the same, but some- 

 times as many as four or five would be seen about a single thistle. 

 On the succeeding day (September 19) it was noticed that the but- 

 terflies had to a certain extent become gregarious. They were no 

 longer evenly distributed over the fields, but were to be found in 

 more or less widely separated; areas where, for instance, from half 

 a dozen to a dozen would be feeding on the goldenrod in a space 

 10 or 15 feet square, or from 6 to 10 would be seated on the heads 

 of a large thistle, with others on the near-by goldenrod. Though in 

 places 2 or 3 could be taken in a single sweep of the net, there would 

 be none on the goldenrod for 200 feet or more between the little com- 

 panies of well-separated individuals. Although their component in- 

 dividuals might be several feet apart, these little companies were 

 constantly on the move. But they moved as individual butterflies, 

 not as if the company were a unit. A butterfly would start up and 

 fly for perhaps 100 yards, settling on a goldenrod some distance — 5 

 or 10 feet perhaps — from another individual. Then a third would 

 settle near these two, and soon another little company would be 

 formed. Sometimes on being frightened a whole company would 

 move off in the same general direction, but the individuals always 

 seemed to scatter more or less. The procedure seemed to be, for the 

 most part at least, a continual forming and breaking up of small and 

 loose aggregations of independent individuals. But it was noticeable 

 that the majority of the butterflies would keep within a certain area 

 which was continually shifting north or south along the belt where 

 the goldenrod was most luxuriant. 



Though most of the butterflies flew only a foot or so above the 

 tops of the goldenrod, one was seen (Hugh U. Clark) to drop from 

 a great height, being first observed on the downward path about 100 

 feet above the ground. 



On the following day (September 20) further changes had taken 

 place. As we reached the field, at 2.30 in the afternoon, we saw 

 within a few minutes' time four butterflies from 20 to 100 feet or 

 more above the ground headed southward but drifting westward 

 toward Great Falls before a moderate wind. This was in the western 

 end of the field where the goldenrod is stunted and scattered and is 



