FAIRY FLIES AND THEIR HOSTS. 



47 



discovering the eggs of the two first named is increased from the fact 

 that they are deposited in the stems and leaves of grasses and various 

 plants, and it is only by careful dissection of such that the host eggs 

 are found, sometimes embedded in the stem or actually between the 

 upper and lower skins of a blade of grass ! or in the soft pith of a 

 rush (J uncus), where a large Frog-hopper occasionally lays a row 

 of twelve to fifteen flask-shaped eggs, each of which may contain 

 four to five larvae of an Anagrus. In another species of Frog- 

 hopper, another species of Anagrus lays but a single egg, the trans- 

 formation of which from egg to perfect Fairy Fly I have frequently 

 observed. 



I first noticed that some of the rushes were covered, at the lower 

 part, with rows of minute punctures, which, on being examined under 

 the microscope, revealed the heads of from six to nine eggs. On 

 stripping the outer covering from the rush, I found embedded in the 

 white pith these flask-shaped eggs, fitting close together and white 

 in colour — the head ends converging, so that the tips could just be 

 seen through the hole in the rush. Some years elapsed before I 

 discovered one September a Frog-hopper, Liburnia (Fig. ii), in the 

 act of laying these delicate eggs ; and when lying flat down on my 

 stomach, with my head buried among the grass and rushes, I saw a 

 Fairy Fly {Anagrus) walking up a rush, and as it walked it beat its 

 clubbed antenna against the rush, until they came into contact with 

 the heads of the Frog-hopper's eggs, when the Fairy Fly showed 

 great excitement as she rapidly examined and felt the eggs 

 with her antennae. In less than half a minute she protruded her 

 ovipositor and pressed the tip against one of the eggs, and after 

 straining a good deal the delicate auger went through and was 

 driven half its length into the egg. For eleven minutes there was no 

 apparent movement of the body, and then the ovipositor was 

 slowly withdrawn and another egg attacked in the same manner. 

 After this I observed the Fairy Fly go from egg to egg. In one 

 qase, immediately on the ovipositor being withdrawn, I isolated and 

 cut open the host egg and succeeded in finding the egg of the 

 Fairy Fly (Fig. 12), and in each one so dissected from the host egg I 

 could with a high power see the nucleus and nucleolus. As the host 

 egg was semitransparent, I could just detect the larva of the Fairy Fly 

 moving about in the fluid matter. In March the larva appeared 

 full-grown and about filled the host egg, the mass of cells being of 

 a red colour. I then dissected several and made enlarged drawings. 

 (Fig. 13.) The mouth was a sucking one — with two simple curved organs 

 which I diagnosed as the mandibles. At each side of the mouth 

 were soft fleshy protuberances, not unHke short stumps, which the 

 larva used as forelegs. The tail was the narrowest part of the body, 

 with a slight projection on the ventral side. 



When full-grown, the body of the larva appeared full of reddish 

 cells. In a few weeks these cells began to move slowiy and crowd 

 together, and as I watched them day by day under the microscope, 



