234 THE AMERICAN NATURALIST. [VoL. XXXIII. 
much lessened. Since the adult salamanders are extremely 
slippery and often very rapid in their movements, it is advan- 
tageous to keep in one hand a small net of cheese cloth, having 
an aperture about six inches across, and with a very short 
handle. 
The larvæ of both species are to be sought for in the water, 
and may be seen lying upon the bottom of the quiet pools, 
especially those with a fine gravel bottom. Even at this stage 
they are fond of concealment, and if there are small stones, 
fallen leaves, or other objects in these miniature basins, they 
should be removed and the water allowed to settle. The larvee 
may be easily captured by means of the net. This should be 
laid upon the bottom and the larva driven into it by approach- 
ing it from behind with the hand or a small stick. The great 
majority of the larvz collected in this way are those of Spe- 
lerpes, as it remains much longer in the larval state, but the 
very similar larve of Desmognathus occur in similar places 
and are very difficult to distinguish from the others. (See 
_ below.) 
Adults. — As the two species belong to different subfami- 
lies, it would seem an easy matter to distinguish the adults, but 
unfortunately the most distinctive characters are skeletal, and 
the external feature, such as color, number of costal folds, etc., 
although noticeably different in extreme or typical specimens, 
show so many gradations and intermediate forms that the deter- 
mination is in many cases extremely difficult. Both species 
are dark above, marbled along the sides, and without pigment 
ventrally. Both species possess a broad dorsal stripe with 
crenulate edges. This stripe in typical specimen of D. fusca is 
very dark brown, so that in living specimens it merges almost 
indistinguishably into the dark slate color which limits it later- 
ally. In the other species the dorsal stripe is usually light 
brown or fawn color, lighter at its outer edges and bordered by 
a very dark brown stripe, hence ‘‘dz/ineatus.” It is, however, 
quite usual to find specimens of D. fusca with a light rufous 
dorsal band, set off very conspicuously from the slate-color at 
its outer edges ; while in many specimens of S. bilineatus the 
dorsal band is quite dark, without lighter edges, thus blending 
