1892.] Comparative Physiology of Respiration. 825 
what part of the respiratory process issubserved by the aqueous 
and what by the aerial part of the respiration? So far as I am 
aware this problem had not been previously considered. It 
was apparently assumed that there were in these fortunate 
animals two independent mechanisms, both doing precisely 
the same kind of work, that is, each serving to supply the 
blood with oxygen and to relieve it of carbon dioxide as 
though the other was absent. That was a natural inference, 
for with many forms the respiration is wholly aquatic, all the 
oxygen employed being taken from the water and all the car- 
bon dioxide excreted into it. On the other hand, in the exclu- 
sively air breathing animals, as birds and mammals, the 
respiration is exclusively aerial. 
This natural supposition was followed in the first investiga- 
tions on the respiration of the soft-shelled turtles, and while 
it was proved with incontestible certainty that they take oxy- 
gen from the water like an ordinary fish, that is, have a true 
aquatic in addition to their aerial respiration ; there was alto- 
gether too much carbon dioxide in the water to be accounted 
for by the oxygen taken from it, Furthermore, upon analyz- 
ing the air from the lungs of a turtle that had been submerged 
some time the oxygen had nearly all disappeared and but 
very little carbon dioxide was found in its place, while, as 
compared with human respiration for example, a quantity of 
carbon dioxide nearly as great as that of the oxygen which 
had disappeared, should have been returned to the lungs. 
Likewise in Prof. Wilder’s experiments with Amia (7), to use 
his own words: “ Rather more than one pér cent. of carbon 
dioxide is found in the normal breath of the Amia, but much 
more of the oxygen has disappeared than can be accounted 
for by the amount of carbon dioxide.” Everything thus 
appeared anomalous in this mixed respiration, and instead of 
a clear, consistent and intelligible understanding of it there 
seemed only confusion and ambiguity. Truly these seemed 
like “ bad experiments.” 
It became perfectly evident that the first step necessary in 
clearing the obscurity was to separate completely the two 
respiratory processes, to see exactly the contribution of each 
