HABITS OF THE GIANT SALAMANDER. 527 



practically one of swallowing air. When it becomes necessary for the 

 hellbender to take in a fresh supply of air, it swims towards the sur- 

 face of the water until the tip of the snout is exposed ; then, by swelling 

 downward the skin of the floor of the mouth, a huge mouthful of air 

 is drawn in, and the animal, now arching upward its neck and back, 

 slowly sinks to the bottom, the swallowed air passing back to the lungs 

 either by its own buoyancy, the lungs now being higher than the throat, 

 or by a sort of peristaltic action of the throat. As the animal sinks, 

 a considerable portion of the air that was taken into the mouth escapes 

 through the nostrils or through the gill openings. After reaching the 

 bottom, the animal frequently retains, for some time, the strongly 

 arched position, as though the region of the lungs were buoyed up by 

 the air that had just been taken in. The air is gotten rid of partly 

 by a quick expiration when the animal comes to the surface, and partly 

 by an occasional bubble sent up from the bottom. If the animal be 

 alarmed as by a sudden approach to the tank in which it is contained, 

 it frequently sets free one or more large bubbles of air, perhaps by the 

 involuntary relaxation of certain muscles, perhaps by a voluntary ex- 

 pulsion of superfluous air, to enable it better to escape the supposed 

 danger. The length of time that the hellbender remains under water 

 seems, in captivity at least, to be quite variable. I made a series of 

 observations on three individuals, a very large one, a very small one 

 and a medium-sized one, and the average interval between inspirations 

 was about fifteen minutes. The longest time that any individual was 

 actually observed to remain below the surface was forty-three minutes. 

 In motion the hellbender is usually slow and awkward. On a 

 smooth dry surface, as the top of a table, it is almost helpless, because 

 the slime secreted by the skin soon dries and becomes so sticky as almost 

 to prevent motion. Even in its native element its motions are usually 

 awkward, though when swimming rapidly this is not apparent. In 

 crawling over the bottom, the diagonally opposite legs very nearly 

 'keep step,' i. e., the left front leg and the right hind leg move forward 

 at the same time. In active swimming the tail is the most effective 

 organ, as in the alligators. If an individual, which is lying quietly on 

 the bottom, be watched carefully it will frequently be noticed that it 

 has a slight rocking or swaying motion, caused by the alternate straight- 

 ening or relaxing of the legs, similar, perhaps, to the swaying motion 

 of elephants. In captivity they usually congregate in the darkest 

 portions of their tank, crawling under boards or stones if these be 

 present, or under each other if there be no better hiding place. If 

 not lifted above the surface of the water, they may be handled with 

 scarcely a struggle, but if taken from the water they struggle to regain 

 their native element. As a rule the smaller specimens were more active 

 than the larger. 



