No. 438.] 



AMBL \ STOMA TIGRINl T M. 



395 



1 )ho>is m !ar\ ;r o >nhned st ru tly to aquatic respiration by nettings, 

 so arranged as to prevent the larvae from rising to the surface 

 for air. But the surprising difficulty was at once encountered 

 that these larva' are, all of them, air breathers from a very early 

 stage, and usually cannot, or will not, endure for any length of 

 time complete exclusion from air-taking at the surface of the 

 water. W ith larva; of but two centimeters in length, when the 

 development of lungs is but slight, confinement below the surface 

 proved fatal. Even with but two specimens in a large aquarium 

 jar, where they had grown from the egg and thriven in the most 

 natural manner possible, the larva' repeatedly bored through the 

 netting at night to reach the surface of the water ; upon the 

 netting being doubled, one larva still penetrated it, while the 

 other was dead beneath it. This experiment was repeated with 

 many variations in method, but little divergence in result. 

 Larvae with very large branchiae, taken from deep, cool, clear 

 water, and introduced singly into a large aquarium freshly filled 

 with the very tap-water in which they had grown would some- 

 times die of asphyxia in a few hours if prevented from reaching 

 the surface. Such results are very striking. " These animals," 

 said an observer of my experiments, " have gills by the whole- 

 sale ; but they seem to be mainly for ornament." Not only did 

 the largest gilled specimens succumb when confined beneath the 



they showed no greater' ability to live under water than the 

 specimens with the smallest branchias that could be chosen. 

 Indeed, when, after man}' trials, we finally secured two specimens 

 that were indifferent or nearly indifferent to their confinement 

 under running water, neither of them were large-gilled forms, 



gilled specimen. Even specimens in the earlier stages of meta- 

 morphosis show little less resistance under this treatment ; and 



Hand in hand with this inability to live as gill-breathers goes 

 the complete ability to live as lung-breathers, even at very early- 

 stages. Seldom, if ever, have my larvae died of bad or poorly 

 oxygenated water. Over three hundred larvae, of perhaps two 

 to four centimetres, were, in one instance, transferred directly 



