i8 
rsiC/zE 
[F'ebruary 
however, can exist, temporarily if necessary, on a diet of though it does 
not thrive on this, as lir. Dupree proved when the writer was with him. Anoph- 
eles will bite each other viciously, tearing off the hairs so frequently that perfect 
specimens of the larvae are to be obtained only by isolation ; will devour skins, 
occasionally kill other larv'ae and always eat any which die. i nino/aenia is sup- 
posed to have the same habits as Anopheles, save that, whereas Anopheles skims 
the surface, C'rano'aenUi feeds just below. From the shape of the mouth parts, 
however, the writer would judge the latter to be distinctive in its method, a vege- 
tarian and also feeding on minute Crustacea and the like. 
In the non-insectivorous groups are two distinct types of mandible and max- 
illa. The facts on hand suggest that these are associated with two types of 
breathing tubes and antennae, as well as with different habits of feeding. The 
author hopes to make further observations when opportunity offers, not having 
hitherto paid special attention to this point. 
The group with long breathing tubes (Sf. n trains, C. pplens, C. territans, C. 
C. restnans, C. salinarius ), feed with the tubes almost continually at the surface, 
sailing about by the movement of their mouth-brushes, and leaving the surface 
only when disturbed. Their food evidently floats freely, and is drawn in by the 
currents set up by the brushes. In these forms the antennae have very large and 
plumose tufts; the marginal comb on the mandible is reduced to papillae bearing 
a few’ hairs and moved over the top somewhat onto the opposite side from the 
lateral comb. The maxillae are of a pointed, sub-conical form, with the terminal 
hairs freely movable and very long, many of tliem as long as, or longer than, the 
bodv of the maxilla, with several long-plumose hairs in the tuft, which is mostly 
composed of single hairs. The plumose hairs are probablv sensory, w’hile the 
function of the simple hairs is probably only to brush in food. 
C. diuolo’ forms an apparent exception, having this type of maxilla and 
mandible, but w’hen one considers that, although he feeds at the bottom, he lies 
on his back most of the time, and feeds on floating matter, his not feeding near 
the surface makes no great difference. 
C. jamaicencis, ‘7 mnsica, and 'J. vm ipes also belong to this type, though the 
latter tw'o can stay at the bottom a long time if disturbed. 
The bottom feeders, as far as known, have straight, stubby antennae, gen- 
erally bare of spines, with one hair or a very few simple hairs in the tuft. The 
mandible has the marginal comb well developed, the maxilla is a flattened cone, 
the terminal tuft of stiff', straight hair.s, which are rather short, is evidently for 
use in brushing over the debris at the bottom, w'here the very long, slender hairs 
would be a disadvantage. C’. dnpreei, with its very short antennae, and its bottom 
feeding habits, is a very good example. C. Irise/ialns lives in hollow trees, and 
