SHORT STUDIES OF TURTLES. 259 



remained voluntarily as long beneath the surface, as any 

 of the others. In fact, in its movements and degree of 

 activity, it most closely resembled the mud-turtle, which 

 is so essentially an aquatic species. Referring to the 

 respiration of turtles, Professor Agassiz remarks : " In 

 mud and soft-shelled turtles, the lungs being much re- 

 duced in size and importance, by far the greater part of 

 the respiration must be performed by the skin of the 

 whole body, which is much thinner in these families than 

 in other turtles; while, on the contrary, in ... the 

 Testudo (box-tortoise) the powers of respiration are no 

 doubt performed entirely by the lungs " (loc. cit., p. 276). 

 In the case of the Muhlenberg turtle the skin is not thick, 

 but, being greatly wrinkled, appears so ; and the lungs 

 are, I believe, of about the same size as those of the 

 strictly aquatic species. Certainly they are very little 

 larger, and correspond more nearly to those of the snap- 

 per than to the terrestrial box-tortoise. Therefore, I am 

 disposed to consider, and feel little hesitancy in asserting, 

 that the Muhlenberg turtle is essentially an aquatic spe- 

 cies. 



A few words concerning the vocal powers of this 

 turtle. In brief articles on the habits of our turtles, 

 published in years past, I have mentioned the voice of 

 the common snapping-turtle. Since then, I have been 

 fortunate enough to hear this same turtle make similar 

 sounds in its native haunts. In every case they were 

 made at night ; but my opportunities were such that I 

 am quite positive that the sound heard and the turtle 

 seen were correctly associated. 



In the often - mentioned monograph by Professor 

 Agassiz, that author remarks : " Turtles have a voice. 

 Though I have myself made this observation only in a 

 few species, namely, in Emys elegans, serrata, picta, and 



