MA Y-FL Y FISHING. 359 



with his elongated foretarsi (whose articulation with the tibia is 

 so constructed as to admit of supination of the tarsus), he bends 

 the extremity of his body forwards over his back, grasps with his 

 forceps the hinder part of her seventh ventral segment, and with 

 his outer caudal setae embraces her sixth segment. These two 

 setae exhibit near their origin a strongly marked articulation, where 

 they can be deflected abruptly so as to lie forwards over the back 

 of the female, parallel with one another between her wings,, 

 Meanwhile the couple gradually sink, the female not being quite 

 able to support herself and mate, and by the time they reach the 

 ground, if not before, their connection is usually terminated.' 



The fecundated female, after resting awhile, repairs to the water 

 and, hovering over it, just touches the surface from time to time 

 as she drops part of the eggs. 



The only purpose for which they seem to have existed in 

 the winged state, viz., that of perpetuating the species, having 

 been accomplished, both sexes fall almost lifeless on the water, 

 with their wings extended and lying flat, and at this period of 

 their brief existence are usually designated by anglers the Spent 

 Gnat. 



The immature May-flies at the earlier stages being burrowing 

 larvae living in the mud ( l larves fouisseusesj as Pictet styles 

 them), do not to any great extent serve as food for the fish. A 

 certain number are occasionally found during the spring in the 

 stomachs of trout, but it is probable that when an autopsy 

 reveals the presence of any considerable number of these larvae, 

 it is due to some disturbance of the mud of the river having 

 set the larvae adrift, and, naturally, a hungry trout finding a 

 quantity of palatable food such as this within his reach would, 

 if possible, gorge himself with it. I know, from undoubted 

 evidence, that from one fish in the Kennet more than one 

 hundred May-fly larvae were taken in spring, but considering the 

 enormous quantity of May-fly present in this river, it is not sur- 

 prising that after a flood or during weed cutting a fish should 

 find hundreds of larvae helplessly carried along by the stream. 



The first stage at which the trout can get the chance of 



