458 REPORT OF THE BUREAU OF ANIMAL INDUSTRY. 



only very swift, but impeded by trees, brush, stumps, or cane. If the 

 water was seething and foaming, in its onward course, among these 

 obstructions, so much the more attractive it seemed to prove as a 

 place of deposit. By swinging a dugout, myself therein, into the 

 midst of these boiling rapids, and holding it there, I was able to 

 watch the female gnats in their work of oviposition. 



The eggs are placed just above the water's edge, upon almost any- 

 thing situated in the midst of the current, vertical objects seemingly 

 preferred, but, nevertheless, many eggs were placed on the sides of 

 my dugout. Very few eggs are deposited during the forenoon, and 

 as they are dropped about without any system in the arrangement, 

 adhering to the object upon which they are placed, if the flood be 

 falling, there will appear for each day a ring or band of eggs, and 

 these bands, at a distance of a few yards away, have the appearance 

 of rings of scum which has collected and dried. So far as observed, 

 eggs hatch within a few hours, and the young larvae at once make 

 their way into the stream. Eggs have, however, been kept several 

 days, and yet hatched at once on being placed in water. 



THE LARV^l. 



On first emerging from the eggs these are very minute, and it would 

 seem that to trust their tiny bodies to the mercy of a strong current 

 would be to commit suicide, particularly as full-grown larvae can not 

 swim about freely, as most other aquatic larvse do. Yet the little things 

 are not easily killed, and seem to be as able to withstand the rough 

 usage of their watery home as when much more fully matured. So 

 far as known the only means of locomotion which these larvse possess 

 is by crawling about on submerged objects with a slow gait, and 

 letting themselves out from such objects by the aid of a tiny, thread- 

 like spider-web, which they attach at one end and dangle in the midst 

 of the foaming current at the other, sometimes as many as half a 

 dozen on a single thread. 



The food of these larvse probably consists almost exclusively of mi- 

 nute animals, which infest the waters of these inland streams in vast 

 numbers, and this is doubtless one reason why these larvse congregate 

 only in places where the current is swift, as a fresh supply of food 

 is thereby being continually brought within their reach. One thing 

 is certain, the larvae, if confined in still water, perish in a few hours. 

 This, however, might be in part due to a lack of oxygen in still 

 water. 



The larvse of most of these Simulium live in the streams nearly an 

 entire year before changing to pupse. There is no indication of more 

 than one brood of adults each year, unless those found in August 

 should prove to belong to a species differing in this respect from all 

 other known southern forms or species. 



THE PUP^B. 



In spring, on becoming full grown, the larvse spin a tough brown 

 cocoon, with the upper end open (see Report of the Commissioner of 

 Agriculture, 1886, Plate VII, Entomologist's Report), within which 

 they transform to pupae, and thence, in about ten days, to adult 

 gnats. 



Unlike similar insects whose habitat is much farther north, the 

 larvse do not seek secluded places, out of the swift current, wherein 



