156 Mr. C. B. Williams' Eecorth of 



great to ]nake it qiiito impossiblo for nio to tell to what 

 family or even to what order the insect belonged nntil 

 one was captured, and over one hour's continuous effort to 

 capture specimens, passing sometimes almost every second, 

 only resulted in the capture of five individuals. 



Prof. Poulton informs me that all these five specimens 

 were females. 



This concludes my own observations on the insect, as it 

 was not seen again during my stay in the district. 



To summarise :■ — The butterfly flew in large ruimbers, 

 at a great speed, in various directions, but chiefly from 

 north or north-west to south or south-east and vice 

 versa, in the late afternoon and evening * of cloudy, close, 

 still days during April and May with a very slight indication 

 of a change in direction as the evening progressed. In 

 addition all the specimens captured were females. 



It may be maintained that a flight of this nature is not 

 a true migration, and it is possible that this is so. At the- 

 same time it is a flight which differs distinctly from the 

 normal fluttering of the insect lound flowers, and as such 

 would be worthy of record. There are, however, certain 

 known observations with regard to this insect which make 

 it probable that this ha})it, if not always truly migratory, 

 may sometimes at least lead to a migration. 



Ccdpodes elhlins feeds in its larval stage on the leaves of 

 Canna. Its distribution is from S. Carolina to the Argen- 

 tine Republic, bnt it appears to be unable to survive the 

 winter much further north than Florida. On occasions, 

 however, the insect appears in large numbers as far north 

 as the district of Colombia, and rarely even as far as Long 

 Island, New York. 



F.H.Chittenden (U.S. Dept.Agric. Bur. Ent., Circular 145 

 (1912), pp. 1-2) gives some particulars of an invasion in 

 September 1904, when thousands of caterpillars were 

 destroying cannas in Alabama, and in the following year, 

 1905, when they reached as far as Washington D.C. He 

 writes (p. 2) : " The cause of this invasion was apparently 

 that the summer of 1905 was an unusually hot one. Heat 

 favours an increase of insects of this type. Prevailing 

 southerly winds, however, were probably more potent 



* Chittenden (U.S. Dcpt. Agric. Bur. Ent., Circular 145, p. 7) 

 says, " Wittfield adds that one of the favourite, times for flight of 

 tlie butterfly in fair weather is after sundown," but 1 have been 

 imablo to trace the original reference. 



