10 



present inaccessible portion of tlie raft, wliicli is more likely to bear 

 cocoons in (jnantity than the floating material at hi^li water.* 



The relation of Bayou Pierre to tlie Ked River is such, as shown in 

 the report of Captaiu Bergiand contained in the Annual Report of the 

 Chief of Engiueers, U. S. Army, Part II, 1885, Appendix U, pages 

 1187-1493, that in times of high water three-fourths of the discharge 

 of the lied River is through this bayou, and in times of low water but 

 one-fourth. Before the formation of the raft, this very great augmenta- 

 tion of the bayou in high water had no ill effect, but now the water, 

 checked by the raft, floods every spring much of the adjacent low-land, 

 thus furnishing additional foothold for larva?, and possibly also driving 

 the adults in larger swarms to the higher land. 



During the summer months the water is confined by moderately high 

 clean banks and is free from drift, except where such material is held 

 by the raft. This would indicate that the raft is largely responsible 

 for the abundance of the gnats in that locality. 



The sumller streams in the neighborhood dry up in the course of the 

 summer, and hence could not breed gnats; however, a number of the 

 princii)al ones were carefully searched for cocoons orlarva^ without the 

 discoxery of any evidence of them. 



The Buffalo Gnat was reported to be quite abundant on the Sabine 

 River; ami it also occurs in less numbers on the Red and Washita 

 Rivers. 



The raft in Bayou Pierre originated in the attempt of the United 

 States Government (in 1872-'73) to close Tone's Bayou, which connects 

 Bayou Pierre with the Red River, and to confine the water of Red River 

 to its own channel. A large raft which was being removed from above 

 Shreveport was run into Tone's Bayou and the attemi)t made to letaiii 

 it there by means of a boom. This raft and also a second one formed 

 later were entirely swept away by floods and carried into Bayou Pierre, 

 where they are at present lodged. As described by Captain Bergiaud 

 (1 c.) the rait " extends 5.3 miles above and 2.8 miles below Red Bluff 

 at the mouth of Wallace Lake. The upper portion is fragmentary, of 

 recent formation and loose structure, occupying in the aggregate one- 

 fifth of the area of the water surface. That below is nearly continuous 

 and gradually becomes denser until at its lower end it becomes solid- 

 ified." This lower poition of the raft has now become almost entirely 

 solidified by the massing of the logs and the accumulation of debris, 

 and trees and shrubs are now growing upon it. 



*Mr. Frier.son subsequeutly collected and forwarded to tbo Department a con, 

 siderable qtiautity of material — cocoons, larvaj, etc., from tliis place, concerning 

 which wo (jnote briclly from his letter of May 3 as follows: 



"The water has fallen about 2 feet below its level when the gnats were hatching 

 out. * * * The current is very swift, » * * mid j found that every over- 

 hanging tree, logs sticking out of the water, and the millions of roots on the liank 

 were litmally plastered over with the cocoons for the distance of 2 feet above and 

 below the water." 



