334 NEW YORK STATE MUSEUM 



quently in the little stream flowing through Cot glen, in Sis 

 Mile creek^ and in Caseadilla creek; and have also been col- 

 lected bj Mr A. D. MacGillivra}' in a brook near Axton N. Y. 

 During the early part of May the larvae are still quite small, 

 the smallest found measured 2.5mm in length, and were scat- 

 tered over the smooth rock bed of the stream where the water 

 is swift, but only about 1 inch in depth. If removed from the 

 brook and placed in vials or still water, they soon die, usually 

 within a few hours. 



The larva is a curious black creature, flattened, its length 

 being about two and one half times its breadth at widest part, 

 each of the four intermediate segments separated from each other 

 and from the cephalic and anal portion by deep constrictions^ 

 thus dividing it into six distinct parts. Kellogg says (in the 

 paper just quoted) that the anterior, apparently single segment is 

 composed of the fused head and three thoracic segments, while 

 the most posterior part is composed of the last two abdominal 

 segments, the intervening parts representing each a single 

 abdominal segment. The larva is footless, but each body part 

 bears a pair of small unsegmented, pointed projections, situated 

 on the ventral aspect of the lateral margins. The organs of 

 locomotion consist of six suckers, one of which lies on the 

 median ventral aspect of each body part; thus there is but one 

 sucker for the combined head and thorax, and but one for the 

 last two abdominal segments. By means of these suckers^ the 

 larva clings to the rock bed of the stream. The larva occasion- 

 ally moves about on the smooth surface of the rock, from the 

 necessity of getting farther into the stream as the water les- 

 sens in quantity, and perhaps also, for seeking its food — the 

 diatoms on the surface of the rock. The structure of the sucker 

 is well described by Kellogg {loc. cit.). The larvae breathe by 

 means of small tufts of short thick tracheal gills, of which 

 there is a pair on the ventral surface of each of the first to 

 the fifth abdominal segments. On the last segment there are 

 two pairs of much larger, thicker, fingerlike processes, perhaps 

 also tracheal gills. The writer collected during May many liv- 



