26 



more detail can be made out in them than in most larvae, on account 

 of the perfect transparency of this insect. I have seen them in many 

 of the unspecialized species, and probably they exist in all but are too 

 small to be recognized amid the other tissues ; but they reach their 

 largest development in a small larva (No. 17), which is interesting in 

 more than one respect. I will briefly describe the habits of this larva, 

 as they possibly have some connection with the development of its 

 "fat-sacs." The larva is found on smooth slabs of rock near waterfalls 

 and splashing streams, not in the water but in such positions that the 

 spray from the fall continually keeps the slab moist. Such stones may 

 be found covered with larvae wriggling about in all directions and 

 feeding on the thin layer of slime-like confervae always present on 

 these stones. When full grown (5mm.) they form small mud shells 

 adhering to the rock surface, and change into short thick-set pupae, 

 dark-brown in colour. Figs. 21 and 23 give highly magnified drawings 

 of the fat-sacs dissected out from one of these larvae. On the outside 

 was found a thick brown layer of tissue ; this was easily removed, and 

 within was seen a clear transparent membrane forming a sac. The 

 contents of the sac were of two kinds (i) a fine granular substance, 

 (2) a mass of large transparent spherical bodies, no nuclei being visible. 

 Fig. 22 shows the same organ as seen in the transparent larva (No. 18). 

 Here we see that two transparent threads are given off from the base, 

 and a single one from the anterior end. In C. pusio the following 

 details have been made out concerning its attachment within the body. 

 At the lateral anterior edge of segment 5 is attached a long strand of 

 tissue, which, passing down the side of the body ends abruptly in 

 segment 10. Adhering to the free end, or connected to it by a trans- 

 parent nerve-like thread, is the fat-sac, from the posterior end of which 

 one or more threads pass down for further attachment and keep the 

 sac suspended freely in the body cavity. The strand, composed of an 

 elastic sheath without and a mass of small spherules or fat-drops within, 

 is green in colour, and capable of considerable longitudinal contraction 

 whereby the sac is drawn up as far as segment 9. At the junction of 

 the segment the strand is reduced almost to a thread, is transparent, 

 and devoid of spherules. 



