THE SWIMMING ICHNEUMON FLY. 259 



transparent larvae of corethra, and many other species 

 of diptera. But though most of the great orders are 

 more or less richly represented, no aquatic species of 

 hymenoptera or orthoptera had, till now, been dis- 

 covered. 



' ' The female of Agriotypus armatus, indeed, has 

 been observed to descend the sides of rocks in the 

 Clyde to a considerable depth, and to remain there 

 several minutes. Even this species, however, cannot 

 be considered as truly aquatic, or as being entirely at 

 home in the water, since there is no evidence that it 

 knows how to swim. Mr. Smith also reminds me that 

 Smiera (Chalcis) femorata and S. sispes are said to be 

 parasites on the aquatic larvae of Stratiomys, in which 

 case the perfect insect probably seeks its prey under 

 water ; but I am not aware that it has been actually 

 observed to do so. 



" The species of Hymenoptera have been estimated 

 at about 30,000 in number, and Mr. F. Smith (than 

 whom there is no better authority) informs me that 

 about 12,000 have been already described, 3,500 of 

 which live in Great Britain. From the interest attach- 

 ing in mr.ny cases to their habits, from their marvel- 

 lous instincts, and their curious relations to other 

 animals (no order presenting more parasites and vic- 

 tims of parasites), the peculiarities of Hymenoptera 

 have received more attention than those of almost any 

 other group of invertebrata. 



" Great, therefore, was my astonishment, on the 

 occasion to which I allude, when I saw in the water a 

 small hymenopterous insect, evidently quite at its ease, 

 and actually swimming by means, of. its wings. At 



