508 REPORT OF THE COMMISSIONER OF AGRICULTURE. 



dried up for a long time, an inundation resurrects vast numbers of 

 tliem, and brings them furthermore within reach of the larva?. These, 

 however, are not active during the heat of summer, and an inunda- 

 tion at that time Avill not affect them at jj.11; but if it should take 

 place early in spring, this additional source of food would soon mature 

 vast numbers otherwise doomed to die. 



Pupa and Cocoon. — As soon as the larva? are fully grown they 

 descend towards the bottom of the water to make their peculiar 

 pouches, and many pupse are found at a depth of 8 to 10 feet below 

 the surface; others much higher up. But in shallow water they may 

 be found clustered one above the other, just above the bottom of the 

 stream, their instinct having evidently taught them to provide for a 

 sudden fall in th6 water. Notwithstanding this, with the water fall- 

 ing in the bayous and larger creeks at the rate of 1 foot per day, 

 many pupse are left high and dry. Those of the Turkey Gnats, which 

 are always found just above the bottom of the smaller perennial 

 creeks, are not thus endangered by a low stage of the water, which 

 rises and falls suddenly with every heavy rain, but remains of uni- 

 form depth at other times. 



In one of the breeding-places of the Southern Buffalo Gnats, at the 

 junction of Crop and Mill Bayous, in Tensas Parish, Louisiana, Mr. 

 Fillion found immense numbers of the dry and empty pouches as late 

 as June 10, 1886; they were attached to vines, trunks of living trees, 

 and leaves retained by the vines. All these pouches were found near 

 the highest point reached by the overflow, forming a zone or belt 

 from 3 to 4 feet in width. On July 15, the current, very swift in June, 

 had almost ceased to be noticeable, and the stream had decreased from 

 a width of 45 feet to that of 20 feet; the Crop Bayou was partly dry, 

 and no obstructions or vines of any kind reached the water, which 

 flowed in clear dry banks. The belt of dry pouches was at the lat- 

 ter date high above the water, the lowest being found some 13 feet 

 above it, while the highest reached to the mark left by the overflow. 



The cocoon or pouch spun by these larvae is conical, grayish or 

 brownish, semi-transparent, and has its upper half squarely cut off; 

 it is fastened to sticks, leaves, or logs. The larva in spinning does 

 not leave its foothold, but running in the center of its work, uses its 

 mouth to spin this snug little house. In it it changes to a pupa, 

 which has its anterior end protruding above the upper rim. These 

 pupse are at first of a light brown color, afterwards changing to a 

 pinkish cast, and, just previous to the hatching of the fly, to black. 

 During the first of the coloration epochs they are attached to the veg- 

 etable substances upon which the pouch has been fastened by the 

 thoracic filaments, by threads about the body, and by the anal ex- 

 tremity; but during the last two the pupa? hang by the short anaL 

 attachment alone to the threads at the bottom of the pouch, and rise 

 more and more out of the pouch, until at last they swing about freely 

 in the current, attached only by the drawn-out threads. 



The pupa itself is distinguished from most other Dipterous pupae 

 by the presence of a tuft of respiratory filaments starting from each 

 side of the thorax. These tufts, as already stated, foreshadowed by 

 two dark spots upon the sides of the thoracic segments in the larva, 

 are composed of a greater or less number of very slender filaments, 

 varying in number in tlie different species of Simulium. Along the 

 posterior margins of each of the third and fourth dorsal segments there 

 are eight minute spines; the tip of the abdomen is also armed with 

 two larger and bent spines or hooks, by which the pupa is secured to 



