as before, substituting Paris green for London purple in both mixtures. 

 Light rain followed the same day, and more on the 9th. On the 10th 

 a scorching of the leaves was somewhat evident, a little more so where 

 the stronger mixture was used, while on the 18th the condition of the 

 foliage was practically the same as on those branches treated with Lon- 

 don purple — if anything, a little less severely injured. There was also a 

 barely perceptible difference in favor of the weaker mixture. Suppos- 

 ing that all the worst injured leaves were rendered practically useless 

 to the tree, the loss of foliage would probably amount to 4 or 5 per cent. 

 There can certainly be no further question of the liability of the cur- 

 culio to poisoning by very moderate amounts of either London purple 

 or Paris green while feeding on the leaves and fruit of peach or plutn; 

 but much additional experiment is needed to test the possibility of pre- 

 venting serious injury to these fruits by this means. The pupal hiber- 

 nation and late appearance of a considerable percentage of the curcu- 

 lios make it possible that sprayings must be several times repeated, and 

 perhaps carried further into the season than is consistent with safety; 

 and the limit of tolerance of these poisons by the peach under ordina- 

 rily trying circumstances has not been clearly ascertained. Further, 

 the observations above reported on the food plants of the curculio make 

 it likely that, in nature, a smaller proportion of the food of these bee- 

 tles comes from the peach or plum than has hitherto seemed probable, 

 and that poisons there applied would kill less certainly. It seems worth 

 while to make the attempt to attract the adult to flowering plants in 

 the orchard other than the peach, with the hope of poisoning it there 

 (especially late in the season) without using these dangerous insecticides 

 on fruits afterwards to be eaten. 



REPORT OF A TRIP TO INVESTIGATE BUFFALO GNATS. 



By C. L. Maelatt, Assistant. 



Washington, D. C, April 22, 1889. 

 Sir: Iu .accordance with your letter of instruction of April 5, 1889, I proceeded to 

 Frierson's Mill, La., and studied, as far as the conditions would permit, the relation 

 of the raft of logs in Bayou Pierre to the injurious abundance of the gnats in that 

 immediate locality. Examination was also made to determine the feasibility of re- 

 moving the raft to prevent the further breeding of the gnats thereon. 



I wish here to express my thanks to Mr. G. A. Frierson and brothers for their kind 

 hospitality, and for the efficient aid rendered by them in the investigation of the raft 

 and bayou. 



Respectfully, 



C. L. Marlatt. 

 Prof. C. V. Pi ley, 



U. S. Entomologist, Washington, I). C. 



As you had surmised would be the case, the Buffalo gnats had already 

 disappeared when I arrived at Frierson's Mill. A few Turkey Gnats 



