nv \KRA IRWIN SMITH. 513 



8th June, when 8.2 mm. long. A single moulting occurred in a number of 

 cases among the larvae taken in April and May. The majority of these were at 

 the 7 mm. stage, but tliere were a few at ti, 8, 9, and 9.5 mm. Whether these 

 leugtiis represent successive instars. there is not yet sufficient evidence to prove. 



All the larvae escape from the old skin in much the same way. Before 

 moulting the skin liecomes dry and withered looking, much softer than when 

 functional, and a lighter colour. The process, which usually occupies about an 

 hour, begins by a splitting of the moult skin, along one side, from the third to the 

 ninth or tenth segment, tile larva slowly moving from side to side, and contract- 

 ing. The skin of the anal region remains intact, and that of the anterior end 

 is split off entire, between the thii'd and fourth segments. Frequently the larva 

 emerges from the posterior portion with its head still enclosed in this anterior 

 "cap," which is subsequently shed. The empty moult skin stands out stiffly, 

 retaining the same size and shape as when the larva is still enclosed in it . 



The new larva' skin is a delicate white or creamy colour, its surface flecked 

 with glistening particles, and showing the typical hexagonal pattern very dis- 

 tinctly. It assumes a brownish tinge only very slowly, and is still light in colour 

 at the end of several weeks. Older larvae show the more normal grey-brown 

 colour, which tones very well with the soil in which they live. The coating of 

 particles of dirt which invests most of them increases this resemblance to their 

 environment. 



Pupation. 



During the period of pupation the larval skin becomes much darker, and as- 

 sumes a dry, rigid appearance, by which the condition is easily recognised, 

 although there is no change in outward fonn . Larval skins of male jiupae mea- 

 sure from 7 to 8 mm., those of females from 10 to 11 mm. For some time be- 

 fore the emerg'enco of the adult fly the pupating larva is quite motionless, and 

 to all appearances dead. One found in this condition on the 13th April, did 

 not emerge until the 30th April, so that the pupal stage occupies at least eighteen 

 days. This is a longer period than is given by .Jusbasehjanz, who states that 

 the pupal stage lasts eleven to thirteen days. 



Larval Ii(ihit.'<. 



The larvae are all extremely sluggisli in their movements. Usually they re- 

 main quite immobile for Hve or ten minutes after being disturbed; then begin 

 slow movements of contraction, and, if lying ventral side uppermost, roll over, 

 and crawl slowly and stiffly along, seeking to take cover beneath the soil, a pro- 

 cess which occupies half an hour or more. On a hard surface, or a layer of 

 soil too thin to burrow into, their rate of progression is of the order of 5 to 10 

 mm. in ten minutes. Larvae conflneil in glass pots in the laboratory are usually 

 found wedged amony' the roots of grass soon after the fresh sods are put in with 

 them, and sometimes adhering to a root by the head capsule. It is evident that 

 their main, if not only, source of nourishment is in the juices of the living plant- 

 But I have never been able to detect any scars, or perforations, or other evidences 

 of injury on the roots, and, even where the larvae are yery numerous, the grass 

 which harbours them shows no ill-eft'ects from their presence. While living 

 normally in soil in which a fair degree of moisture is present, they are able to 

 sustain life in much drier conditions. Individuals left for twenty-four hours or 

 more, without soil, in a dry Petri dish, showed no ill-effects from the experi- 



