first stages of Pericoma and the Brazilian species. 485 
not mentioned in Dr. Miiller's letterpress. The larva of 
P. canescens has, on the sides^ long and stiflf hairs, all 
pointing backwards,, like those which project from the 
dorsal shields and probably intended to " save the 
larva from being swept away by a sudden rash of water/' 
Similar setse and hairs on the sides and at the ends of 
the chitinous transverse plates are visible in the larva of 
M. Ursula, bat they are much longer than those of 
canescens, some of them being as long as the whole 
breadth of the larva. In this respect, the two other 
larvae {M. pilosella and spinosa) come nearer to P. 
canescens in the shortness of their appendages : " their 
abdommal segments/' says Dr. Miiller, " are beset on the 
sides with straight spines '' ( Dornen), the length of 
which is far less than half the breadth of the abdomen ; 
in one of the species M. pilosella (figs. 2, 3, 4) these 
spines are simple, in the other (M. spinosa) they bear 
from four to six and even seven branches. 
The dorsal view of the two larvge (at least in the 
figures) shows a difference in the structure of the 
breathing organs. P. canescens shows, on the prothorax, 
a pair of projecting cylindrical tubes, which are the 
anterior spiracles. In Jf. iirsula such tubes are not 
visible. Tlie anal breathing apparatus in the Maruina- 
larvae is represented by the two types of structure, which 
have already been mentioned above as described and 
figured in the ^' Eutom. Nachrichten.'' 
One of the larvas {M. pilosella) can breathe either 
directly, through a pair of anal spiracles, or, when under 
water, by means of three pairs of branchial appendages 
(tracheal gills), compare Tab. x., fig. 2 ; and also in the 
Ent. Nachr., 1888, the figures A and B. The other larva 
(if. Ursula, Tab. x., fig. 1 L ; and Ent. Nachr., 1888, 
fig. G) has two tracheal trunks inside a tubular elon- 
gation of the last abdominal segment, with a circular 
fringe of hairs around its opening, that enables the larva 
to hold a bubble of air when it is under water. Short 
branchial appendages, apparently only a pair on each 
side, and not three as in M. pilosella, protrude a little 
above the opening of the spiracle tube. 
The apparatus in P. canescens is somewhat different, 
consisting, at the end of the body, of two pairs of anal 
processes, chitinous rods, bearing a fringe of fine 
filaments, which project from the sides and tip '' (Tab. 
