404 Observations on the Natural History 
This reptile feeds on worms, small bivalves, and snails. In 
this he resembles the salamander, but he bears fasting much 
better, being able to live two years and even more without ali- 
ment. When taken from his natural habitation, and exposed 
to the vicissitudes of the season, like other perfect reptiles, he 
hides himself during winter, is inert, and refuses food. 
The proteus does not live long if he is taken out of the water. 
When he becomes dry, he dies more or less quickly, according 
as the season is more or less warm, being less able to sustain life 
under such circumstances than fishes. But if the proteus die 
more speedily when out of water ; in water, on the contrary, he 
lives better than fishes, since, caderis paribus^ he has not such 
frequent need of a renewal of the water as fishes have. When 
placed in a small vessel, in water at the temperature of 63°.5, 
the proteus, like fishes, rises at times to the surface to take in 
air by the mouth. In doing this, he opens his mouth as wide 
as he is able, and again rejects the air very quickly through the 
branchial apertures. In the act of taking in air, and passing it 
through these apertures, he makes a certain noise, not unlike 
that made by a syringe, when a little air insinuates itself with 
the liquid into the tube; but when the animal is removed from 
the water, and then inspires air by the mouth, this noise is not 
heard. In the escape of the air by the branchial apertures, when 
the animal breathes in air, some minute bubbles remain attach- 
ed to the margins of the apertures, or to the roots of the gills ; 
and the gills themselves collapse and fall down against the sides 
of the head. On the contrary, when the air is received into the 
mouth while the animal is in water, it escapes freely through the 
branchial apertures, and rises in bubbles to the surface. 
The necessity of inspiring air from time to time, is more or 
less great, according as the water is more or less impure ; and it 
has a more direct relation to the temperature than to the quan- 
tity of water. It is greatest when the animal is removed from 
the water ; he is then seen to take in air, and reiterate this ope- 
ration ; his breathing becomes weaker, and at the end of two or 
three hours he. ceases to breathe. But if the water of the vessel 
have the temperature of 63°.5, and be also frequently renevved, 
he then has no need of rising to the surface to inspire air, and 
this is still less necessary if the water be in large quantity, or 
