Insect Mir/ralioHs it) Tropical America. 153 



There is no land to the east of Trmidad, and the only 

 possible explanation of this flight seems to be that the 

 butterflies had migrated out to sea northward from the 

 (luiana or Veneziielan coast, and had been blown westward 

 to Trinidad by the prevailing westerly trade wind. 



In a previous paper {he. cif., 1917, p. 162) I mentioned 

 a record of a flock of tlie small white Pierid, Appias mar- 

 garita, flying northward out to sea at Berbice, British 

 Guiana, in 1909. 



It is extremely urdikely that the two records refer to the 

 same flight, but the latter record at least lends consider- 

 able probability to the jiroposod explanation of the former. 



LvrAENIDAE. 



Tmolns heon in Trinidad. 



On the 23rd March, 1919, I visited the famous Pitch 

 Lake at La Brea, Trinidad [south-westeni promontory]. 

 At 4.15 p.m. we noticed a large number of snuill l)hie butter- 

 flies passing in a continuous stream across the open stretches 

 of the " lake," which is laigely devoid of vegetation. 



The migration was watched for at least a quarter of an 

 hour, and it was in full swing both at the beginning and 

 end of this period, so that it may have lasted much longer. 



The butterflies were flying fast and very low, seldom 

 above five feet from the ground, and often less than a foot 

 above it. 



They were flying almost due south, slightly S.8.W. 

 The wind was a fairly strong easterl5^reeze, so that the 

 direction of flight was almost directly across it. 



In successive minutes on a 30-yards front the following 

 numbers were counted, 25, 8, 12, 15, 17. The flight was 

 going on as far as one could see in both directions, so that 

 even in the short period that we were watching many 

 thousands must have passed. 



I had no net at the time, and only four specimens were 

 caught by knocking down with our hats. Two of these 

 were smashed beyond recognition. The other two were 

 forwarded to Mr. W. J. Kaye, who kindly identified them 

 as Tmolus bean Cram., a connnon Trinidad species. Both 

 were males. 



During the period that we watched only three Lycaenids 

 were seen flying in any other direction, and occasionally a 

 yellow butterfly was following the general rush southward. 



