A. W. lU'MER Roberts 131 



perhaps in very mild weather. In a series of tests made in a field known 

 to be infested, none were found nearer the surface than six inches, 

 while several were found in the subsoil, which in this field was at some 

 10-12 inches from the surface. It is possible that larvae at such a depth 

 might move in their burrows nearer to the surface through the influence 

 of rain and mild weather, but it seems likely that little food, if any, 

 would be taken between the time when they went deep into the soil 

 and that when they returned to feed at the surface in spring. Probably 

 therefore in many cases we have a condition of hibernation similar to 

 that obtaining with many lepidopterous and other larvae, though in 

 others the larvae remain in the sod at or near the surface, feeding on 

 the roots. 



As mentioned above, the larvae in the course of their movements 

 through the soil construct burrows, so that in case of necessity they are 

 able to retreat fairly rapidly. These burrow^s impress the observer when 

 a large number of larvae are kept together in the confined space of a 

 pot and the soil is of such a character that they remain undamaged, 

 when it is lifted out in blocks. The burrows are then seen to honeycomb 

 the soil, ramifying up and down and in all directions. 



In the burrow the larva changes its skin, sometimes enlarging the 

 burrow in the process to form a similar cell to that in which pupation 

 takes place. Such cells are also frequently formed for hibernation and 

 the larvae in a sluggish condition may be found in them in winter. 



Mention must also be made of the propensity of Agriotes larvae for 

 animal food alive or dead. Probably in nature little opportunity occurs 

 to most larvae to indulge their tastes in this direction, but in captivity, 

 as is well known, they will readily feed upon one another, even in the 

 presence of vegetable food. Except in case of injury or at the time 

 of ecdysis, the thick chitin of the cuticle is usually sufficient protection. 

 At ecdysis however the larva is helpless both before and immediately 

 after the moult takes place, so that iu the case of larvae kept in a con- 

 fined space the mortality is very great. 



Pupal Stage. 



Pupation takes place in the ground, within a cell prepared before- 

 hand by the larva. So far as my observations extend, the pupa is placed 

 erect in the cell, with the head uppermost. This habit has also been 

 observed by Hyslop(i6) in the case of the American species Agriotes 

 mancus. Say. 



The pupal instar extends over a period of about three weeks, while 



