[ 62 ] 



colour. Dissection of females some time after oviposition shows that 

 all the ova are not mature at hatching but that there is at least a second 

 batch of eggs. Females captured in the open which are slender bodied, 

 i.e. which show signs of having oviposited, have small pear-shaped 

 ovaries occupying the posterior part of the abdominal cavity. The 

 eggs in these may be well-developed and of the typical shape as when 

 mature, but of the pale salmon colour. This suggests that females 

 having oviposited may continue to live and to produce a fresh race 

 of ova. This second lot of ova is not merely mature at hatching but 

 without shells : they increase in size during the adult period. A female, 

 F 22, hatched on the evening of the 20th July, and which had been 

 mated and had oviposited, died on the 26th. Dissection showed that 

 all the black shelled ova had been laid, and that the ovaries were 

 small in size and confined to the hinder part of the abdomen. In this 

 case, however, the individual ova were much smaller and had much 

 less 3'olk than was found in the fly captured in the open and which was 

 therefore presumably older. The question of the length of life of 

 individual crane flies has not so far been settled. 



Experiments were performed to test the degree of stimulus needed 

 for the act of oviposition. Flies which had been mated and were placed 

 upon cotton wool did not oviposit. Also when the wool was placed 

 upon a layer of soil, they still failed to respond. Only in a few cases 

 were eggs laid upon bare soil, whilst amongst herbage they were deposited 

 readily. Although crane flies are known to oviposit usually in grass, 

 it was found that they may do so in standing corn also. 



The flies kept m captivity were not fed. They usually had access 

 to growing grasses, and to soil, both of which were watered. They 

 were observed licking at moist soil and at the wet sides of the glass 

 vessels in which they were confined. 



EXPLANATION OF PLATE XXXVl. 



View of Tijmla palvdosa in coitu. The halteres are not visible 

 owing to their vibratory movements. 



