THE CANADIAN ENTOMOLOGIST. 103 



sometimes the branches of the spines assume a brownish tint, especially 

 on the anterior segments. 



When full grown this larva measures a little over half an inch ; it is 

 nearly cylindrical, tapering slightly towards the hinder segments. 



The head is rather small, nearly globular, pale green with a faint 

 yellowish tinge, and a dark brown dot on each side, and a few very fine 

 short hairs visible only with a strong magnifier. The mandibles are 

 tipped with brown. 



The body above is dark green, thickly set with green tubercles, from 

 which proceed fleshy looking, forked, pale green, hair-like branches, 

 most of them with their branches extending anteriorly and posteriorly 

 On the anterior part of the second segment there is a row of four spines 

 with five branches each, most of the others are forked, but some few of 

 them have three branches each. There are eight spines or tubercles on 

 most of the segments, arranged more or less perfectly in a double trans- 

 verse row. In some specimens the hair-like branches or appendages are 

 black at the tips, and occasionally entirely black from the point of 

 divergence. 



The under surface is similar to the upper ; feet and prolegs green. 



When mature — from the middle to the latter end of June — these larvae 

 penetrate below the surface of the ground, where they construct little oval 

 earthy cocoons, formed by glueing together particles of earth with silky 

 and glutinous matter. These cocoons are toughly made, and may be 

 taken out of the earth in which they are embedded and even handled 

 roughly without much danger of dislodging the larvae. The specimens 

 which we have bred, when examined a week or two after the cocoons 

 were constructed, were still in the larval condition, although somewhat 

 contracted in length. They all dried up and died before changing to 

 pupae, so we are as yet unable to indicate when this change takes place, 

 the appearance of the chrysalis or its duration. As we have not met 

 with more than one brood in the season, it is probable that the larvae 

 remain in the ground for some weeks unchanged, gradually transform to 

 pupae and remain under ground in this condition until early the following 

 spring. 



