numerous whirls of water caused by the rapid current of the stream 

 impinging against them. 



Careful and continued search on April 15 over 2 or 3 miles of the 

 lower portion of the raft, near Lake Cannisnia, resulted in the finding 

 of a few isolated cocoons on logs which were partially upright, and thus 

 projected several feet into the water. Logs so placed, and possible of 

 removal for examination, were not commonly met with. Nearly all of 

 the floating logs extended lengthwise on the surface of the water, being 

 submerged but a few inches, and hence did not afford suitable condi- 

 tions for the cocoons, and if larvae of the buffalo gnat were on the logs 

 their small size prevented their discovery. 



On the day following (April 10) that portion of the raft near Red 

 Bluff was examined, and here again were found excellent breeding 

 places for Simulium larva?, viz, a swift current striking against the 

 logs and rubbish of the raft forming innumerable whirls and eddies, 

 and somewhat better success attended our search here. On submerged 

 branches, twigs, etc., which projected several feet below the surface of 

 the water and which were evidently raised with the floating lower por- 

 tion of the raft, were found large numbers of cocoons (some few of which 

 contained pupa?) and larva?. A few cocoons and larva? were also found 

 attached to water plants growing from the logs. These specimens were 

 found only where the current set strongly against the raft, this causing 

 the riffles known to be necessary for the larval and pupal existence of 

 Simulium species. 



The larva? and pupa? found proved to be largely if not altogether those 

 of the Turkey Gnat {S. merirfionale Riley). Many of the larva? did not 

 exceed l mm in length, were almost hyaline and apparently but recently 

 hatched (?); others were full grown, and spinning cocoons. These 

 larva? were found attached to the smaller branches and twigs which 

 were in nearly every instance already thickly crowded with cocoons. 

 It is probable from the association of the Turkey Gnat larva? with most 

 of the deserted cocoons that the latter had contained the earlier ap- 

 pearing gnats of this species. Some few of the cocoons may have been 

 those of the Buffalo Gnat, as also some of the minuter larva?, but this 

 could not be satisfactorily determined. 



The height of the water prevented any satisfactory examination of 

 the trees and shrubs growing near the bayou, but wherever possible 

 branches or vines so situated and extending into the water were drawn 

 out and examined. Xo evidence of gnats, however, was found. 



Mr. G. A. Frierson has promised to look for cocoons here as soon as 

 practicable. The reported appearance of the gnats coming to the sur- 

 face in such places in quantity as well as about the raft would indicate 

 that the larva? had during the spring flood attached their cocoons to 

 such submerged trees and branches. Examination will also be made 

 at low water for further evidences of the gnats on the lower and at 



