464 EEPOET OF THE .BUKEAU OF ANIMAL INDUSTBY. 



Several well-authenticated instances are on record where individ- 

 uals have been unable to defend themselves against the attacks of 

 hordes of these insects, and were so severely bitten that they died 

 shortly after in great agony. 



THE EFFECT OF OVERFLOWS ON BUFFALO GNATS* 



That the escape of water from the Mississippi River into those 

 streams whose courses lie through alluvial country, and nearly par- 

 allel with itself, exercises a great influence upon the number of gnats, 

 and consequently upon the ravages of this pest, there is abundant 

 evidence. 



Since the year 1881, every invasion of gnats, of which information 

 can be obtained, has been preceded by a rise in the Mississippi River 

 sufficient to affect the streams wherein the insect is known to origi- 

 nate, and during the breeding season of the years 1886 and 1887 the 

 number of these pests has coincided exactly with the rise and fall of 

 the Mississippi during March and April, which months mainly con- 

 stitute the gnat season. 



The overflow escapes from the Mississippi River through defective 

 levees, or at points where no levees exist. One of these localities is 

 between Gaiiies's Landing and Lana Landing, Arkansas. 



Mr. Robert E. Craig, a very intelligent and observing planter at 

 the latter point, has kept a memorandum of the occurrence of gnats 

 since 1881, and states that he has never been troubled with the pest 

 except when the Mississippi overflowed in his neighborhood. 



In Louisiana where, since 1866, theso gnats have caused so much 

 trouble, there is no record of stock being destroyed prior to 1861, 

 although the people were much more liable to have observed and 

 reported an invasion of gnats had it occurred, than at present. Un- 

 til 1861 the levees were kept in good shape, and no overflow-water 

 filled the inland streams as at present. Again, below the mouth of 

 the Red River no serious damage by gnats has been reported, and 

 such wide-spread ravages are there unknown. Yet I am assured by 

 both artillery and cavalry officers that from 1862 to 1865 gnats de- 

 stroyed many horses and mules in the vicinity of Pointe Coupee. 

 Levees destroyed then are now rebuilt. With so much evidence of 

 the effect of overflow we are led to believe that when a complete 

 system of levees shall protect the alluvial country now infested from 

 this influx of water the occurrence of these gnats in such destructive 

 numbers, at least, will cease. Of the truth or fallacy of this theory 

 we shall not long remain ignorant, as during the present year a com- 

 plete levee system will extend from the mouth of the St. Francis, in 

 Arkansas, on the west side, and from Memphis. Tenn. , on the east 

 side of the Mississippi River, southward to the Gulf. 



In proportion to the volume of water added to the normal supply 

 of these inland streams is the rapidity of their current increased, 

 thereby multiplying the number of localities favorable to the devel- 

 opment of gnats. Such localities are influenced (1) by bringing a 

 larger supply of food within reach of the larvae, and (2) starting at 

 low water each foot of rise adds to the number of objects to which theso 

 larvae can anchor and attach themselves while passing the pupal 

 stage, reaching the maximum in this effect during a general inunda- 

 tion, when the~ greatest destruction of stock occurs. 



It would, however, at present, be too much to say that there are 

 not other ways by which overflow affects the number of gnats. 



