MUSEUM OF COMPARATIVE ZOOLOGY. 25 



But if little or no carbon dioxide is eliminated by the air-bladder, 

 what is the purpose of this aerial respiration] The answer must be 

 evident from the amount of oxygen which invariably disappears. The 

 emitted gas contains only from two thirds to one half the amount of 

 oxygen found in an equal volume of atmospheric air. I believe it is 

 therefore definitely and satisfactorily settled by these experiments, that 

 the air-bladder respiration in Lepidosteus is subservient to the oxygenation 

 of the blood. 



There arise two more questions, which I believe to be intimately 

 related to each other. If the air-bladder does not provide for the 

 elimination of a larger amount of carbon dioxide than is stated, how is 

 this product disposed of] And, secondly, What is the relation between 

 branchial and aerial respiration in Lepidosteus, and how has it been 

 brought about ] 



The most probable method of elimination of the carbon dioxide is by 

 means of the gills. To determine the question definitely would require 

 the complete separation of the effects of branchial respiration from those 

 accomplished by other means ; but it seems to me at present extremely 

 doubtful if any method can be devised which will enable one to do this 

 in the case of Lepidosteus. With Cobitis, or any other fish with intes- 

 tinal respiration in which the gas is emitted through the anus, the solu- 

 tion of the problem would be comparatively easy. 



Bat there is already some evidence that the gills in such cases effect 

 an increased amount of elimination. At least the comparisons which 

 Baumert instituted between Tinea and Cyprinus on the one hand, and 

 Cobitis on the other, indicate clearly that there is a more rapid elimi- 

 nation of carbon dioxide through the gills in the latter case than in the 

 former. 



If, then, it can ultimately be shown that in cases of aerial respiration 

 the gills accomplish most of the elimination of carbon dioxide, and that 

 the air-bladder is serviceable principally as a means of securing addi- 

 tional oxygen, something will have been accomplished toward appreciat- 

 ing the natural conditions which must have ted to the imposition of this 



the air-bladder and from the intestines of several Brazilian fishes gave the 

 following results : — 



I. Fishes with intestinal respiration, 



a, emitting gas from the anus, Callichthys asper C. & Val. : 1.5-3.8% carb. diox. 



b, emitting gas from the mouth and gill slits, Eypostomos : 1.5-2.8% carb. diox. 



II. Fishes with air-bladder respiration, Erythrinus tcematus 1 . , o_ 9 a M v, jj ox 

 Spix, or Erythrinus brasiliensis Spix, > 



