140 Prof. J. Wood-Mason on a Viviparous Caddis-fly. 



the insect began to feel the effects of the alcohol there issued 

 fi'om the extremity of its abdomen in a dense cloud innumerable 

 tiny living creatures, which wriggled convulsively in the fluid 

 for some seconds before they died. These tiny creatures, on 

 examination under the microscope, proved to be Trichopterous 

 larvaj possessing all the characters, namely the slender and 

 tapering body, the laterally-expanded and dorsally-humped 

 first abdominal segment, but above all the disproportionately 

 long and slender third pair of legs, of those of typical Lepto- 

 ceridte. They closely resemble the larva that forms the sub- 

 ject of De Geer's pi. xv. fig. 10 (Hist, des Ins. t. ii. pt. i.), 

 which undoubtedly represents the larva of a species of the 

 same family. They measure about '16 millim. in length and 

 about '125 in breadth ; they number no less than 460, 

 according to my native artist, who measured and counted 

 them for me. As is often, if not invariably, the case with 

 Trichopterous larvse of the first stage, no tracheal gills are 

 present, at least none are to be detected. 



No trace of the gelatinous secretion by which the eggs of 

 the oviparous forms are bound together in masses was detect- 

 able either in the body of the mother or amongst the extruded 

 brood. 



The abdomen of the female still retains the distended con- 

 dition it had before parturition, and presents itself as a thin 

 and transparent membranous sac, the walls of which bear 

 both on the dorsal and on the ventral side a longitudinal 

 series of exceedingly short, transverse, brown bands, repre- 

 senting the more firmly chitinized terga and sterna of its con- 

 stituent segments. The four peimltimate of these segments 

 appear to be extended and stretched, both in the longitudinal 

 and in the transverse direction, to the limit of the extensibility 

 of all their interarticular membranes, being separated from 

 one another both above and below and at the sides by long 

 and equal membranous intervals, while the four basal are 

 stretched to little more than half the extent of their mem- 

 branes in any part ; so that the posterior half of the abdomen 

 would seem to be that which gives lodgment to the main 

 mass of the brood-pouch. The abdomen is in fact expanded 

 for the accommodation of the developing brood much more 

 after the fashion of that of the white-ant queen for her eggs 

 than of those of the viviparous Coleoptera of the genera 

 SinrachtJia and Corotoca described by Schiodte. 



The mother insect, which is of a dull golden-brown colour, 

 has the antennae equal to the anterior wings in length and is 

 furnished with a retinaculum; it agrees in all essential particulars 

 with McLachlan's diagnosis of the genus Notanatolira, to 



