CECIDOMYIM. 121 



to live in a single gall. What causes the galls is not 

 easy to say; probably some peculiar irritation produced 

 by the insect. It is probable that the larvae generally 

 do not undergo moulting like most dipterous larvae. 

 Before assuming the pupa state some larvae leave their 

 galls and conceal themselves under ground, in dry leaves 

 and moss or under the bark of trees; while others, on 

 the contrary, complete their transformation within the 

 gall itself. The pupae are generally, but not always, en- 

 closed in the cocoon, which Winnertz says is produced 

 by an exudation or excretion from the larvae and is not 

 spun. Larvae which had fastened themselves to the leaf 

 were surrounded in twenty-four hours by a white halo of 

 thread-like particles like spicules or crystals, the larvae 

 remaining almost wholly motionless. The cocoon is com- 

 pleted within a few days but even then shows no traces 

 of a genuine web. There seems to be no doubt, how- 

 ever, but that they do in some cases actually spin a 

 cocoon, as is so frequently the case among the fungus- 

 gnats. The change to the pupa state is marked by an 

 alteration of color, the anterior segments of the larvae 

 become distended, and the legs, wings, and antennae de- 

 veloping and rapidly attaining their full pupal dimen- 

 sions. 



The period of pupation is very variable, depending 

 upon warmth and moisture. The development is retarded 

 by cold, though they are capable of withstanding a very 

 low degree of temperature. 



One of the most remarkable things in the biology of 

 these or any other insects was discovered by Wagner in 

 i860. According to this scientist, the larvae of certain 

 species belonging to the genus Miastor, and which live 

 under the bark of trees, produce from ovary-like organs 

 a number of eggs which hatch within the abdominal cav- 

 ity of the parent and there remain, feeding upon the 



