markable for the large quantity of air which escapes from its mouth. The source of this air he has not been able to determine. At certain times it approaches the surface of the water, and seems to take in air, but he could not think that so large a quantity as is seen adhering in the form of bubbles to the sides of the gills could have been swallowed, nor could he suppose that it could be secreted by the gills themselves."

Since the exhalation of air from any source is evidently as easily performed below the surface, the periodical ascent of the gars goes far to show that there is likewise an inhalation. But as it was not easy to determine this, on account of the small size of the young gars and the difficulty of handling the older ones, the writer experimented upon another Western Ganoid, the Amia, or "mud-fish," or "dog-fish."

When placed in a tank the Amia kept near the bottom, and seemed to prefer the darker portions. But it came to the surface at pretty regular intervals, emitting one or two large bubbles from the mouth, and, on descending, several smaller ones from the opercular orifice.

The fish was gradually accustomed to having the body gently embraced by the hand about the middle.

Having been thus prepared, the fish was permitted to swim to and fro in the tank, but prevented from rising. It soon became uneasy, and, after a few not very violent efforts to disengage itself, emitted a large bubble of air.

Now, if this emission were all that was necessary we may suppose that it would have remained quiet for another period. On the contrary, after a second or two of repose (perhaps resulting from the habit of being satisfied after the respiratory act), the fish became more and more uneasy, moved rapidly to and fro, turned and twisted and lashed with its tail, and finally, by a violent effort, escaped from the hand. It rose to the surface, and, without emitting any bubble, opened