LIFE HISTORY AND SEASONAL PREVALENCE 151 



the larvae of dragonflies, but one fish will do more than a 

 hundred of these. Tadpoles do not eat them." 



On the other hand, there is a general concensus of 

 evidence that the larvae of dragonflies are most efficient in 

 this respect, and I believe that their universal presence in 

 all collections of water of any size in upper India is the 

 explanation of the fact that we so rarely find gnat larvae 

 in such situations. The only parasite I have ever met with 

 infesting the larvae is the small stalked infusorian repre- 

 sented in the accompanying illustration. I have repeatedly 

 found every larva in a pool simply covered by these para- 

 sites, which lie crowded together in enormous numbers, 

 attaching themselves especially to the softer parts of the 

 integments, such as the angles between the anal tubercles, 

 and the soft membranes between the segments. 



Larvae affected in this way have a peculiar, slimy appear- 

 ance, and seldom appear healthy, though it is difficult to 

 see how these ecto-parasites can be harmful, unless it may 

 be that being, at the very least, greedy mess-mates, they 

 may appropriate to themselves an undue share of the food 

 that would otherwise fall to the share of their hosts. Never- 

 theless, I strongly suspect that they may be the cause of 

 the inexplicable disappearance of larvae, already alluded 

 to, from situations where they were just before present 

 in abundance. 



Since the above went to press, I have received from Dr. 

 J. Cropper, of Chepstow, a slide containing some immature 

 Acari which he found upon Aii. maculipemiis, in Palestine. 



They are of cordate outline, nearly as wide as they are 

 long, and are provided with a formidable suctorial mouth. 

 From their comparatively large size, they must be formid- 

 able parasites to so small an insect, but only three specimens 

 were observed to be infested in this way. They much 

 resemble the hexapod larvae of certain bird-ticks. 



