No. 438.] 



AMBL J 'STOMA TIGRINl r M. 



tion to the terrestrial form. That animals so treated have, in 

 some cases, metamorphosed is certain; but have sufficient con- 

 trol experiments been conducted to eliminate the possibility of 

 other causes than aerial respiration? The writer has repeated 

 these experiments, and does not find that air-breathing, taken by 

 itself alone, is the natural cause of metamorphosis, or even, in all 

 probability, an accelerating condition. I will instance one set 

 of typical experiments. 



Five larvae, 14 to 18 cm. long, placed on mud with but water 

 enough to partially cover them ; refused to eat ; kept quiet ; 

 all began metamorphosis three days later, and finished eight 



Second : Fourteen larvae, same as above, placed in large 

 aquarium with five inches of water; larvae quiet; not fed. 

 Began metamorphosis four to five days later, and completed 



Third : Twenty larvae, same size as above, placed at same 

 time, in aquarium with water just sufficient to cover back fin. 

 Took little food ; frequently excited and active. Fight showed 



All had completed 3 metamorphosis in eleven days. 



Fourth : Twenty larvae, carefully assorted, with one exception, 

 as to size and other characteristics to match the preceding, were 

 placed in an identical situation ; but with a depth of one foot of 

 water. They ate well from the start, the eating habit being 



all, with the exception of the one peculiar, broad-headed speci- 

 men, almost simultaneously with those in experiment three. 



Fifth. Simultaneously with the above, more than twenty 

 larvae, not assorted as to size, and mostly smaller than the pre- 



but on the sixth day many had begun the transition, and it was 

 completed in about the same period as with the foregoing. 



Sixth. Above fifty small larvae from the same source as the 

 above, deemed quite unsuitable for experiment, as they had not 



