CELL SIZE AND MKTAHOLIC ACTIVITY IX AMPHIBIA. 361 



were not quite standardized. Therefore Xo. 8 is of value only 

 in comparison with Xo. 7. Subtracting the two it is found that 

 the 14 hours of aeration resulted in the accumulation of n mg. 

 of CO 2 , or 0.785 mg. per hour. Comparing this with Xo. 6 it is 

 seen that 1.275 m S- P er hour resulted from the presence of 

 water in the respiration jar in the aeration numbered 6. The 

 source of this CO 2 is the dissolved bicarbonate of the water. 

 \\hich gradually liberates carbon dioxide when < ( >_ tree air is 

 bubbled through it. This is not a factor when an animal is in 

 i lie jar, as is shown by the determination recorded above of the 

 amount \ dissolved CX) 2 in the water surrounding a respirin- 

 animal; -uiticient to prevent liberation of the gas from the 

 IMI arbonate. Trials Nos. I, 2, 3 and 5 indicate that this com- 

 bined (')._. e< lines out quite slowly. Xo. 4 represents the 

 degree \ a < curacy which can be expected from the titration. 

 This figure represents 0.8 cc. of X 10 barium hydroxide in 250 

 .ind thus is an error of 0.32 per cent. 



TAHI.I. III. 

 HUSK TKSTS os CARBON I>K>\H.I 



1 1 appears, then, that the carbon dioxide collet tc< 1 in the blank 

 aerations is in>in three sources: (i) that due to the liberation 

 of the bound ' of the water, (2) that caused !> residual ( ' >_. 

 in tlie atmosphere of the respiration chamber at the end of the 

 preliminary aeration, and (3) that resulting from C(J 2 getting 

 through the air washing system before the air enters the respira- 

 tion chamber. Of these only the last is of importance in pro- 

 ducing error, the other- being eliminated by the presence of a 



