CELL SIZE AND METABOLIC ACTIVITY IX AMPHIBIA. 359 



and show relatively large cells. Rana paluslris Xo. 19 was 

 collected from the field and blood smears made immediately, 

 yet this specimen shows the smallest corpuscles measured for 

 this species. On the other hand Rana palustris Xo. 18 shows the 

 largest corpuscles of the species, and was used when in an 

 extreme state of starvation. There seems to be no correlation 

 between cell size and degree of starvation, as judged by the size 

 of the red blood cells. 



B. Carbon Dioxide Determination. 



1. The apparatus used has two inherent defects. These are 

 not i -on-idrred to be of sufficient importance to affect the results 

 for tin relatively large amounts of carbon dioxide measured. 

 They are, first that rubber tubing was used for all connections, 

 .ind M -oml th.it soft glass bottles were used for collecting jars. 

 Rubber tubing has been shown to have a selective absorption for 

 <.irbi.ii dioxide, but this should not be important considering 

 the -hort lengths of tubing used, the rapidity of the air stream, 

 .iml the relatively large amounts of carbon dioxide collect^ 1. 



2. To learn whether the traps to remove carbon dioxide from 

 tin- air before it entered the respiration chamber were taking 

 out all the gas, a gas washing bottle omtaining a carefully 

 measured amount of standard barium hydroxide was placed 

 between the traps and the respiration chambers. 22 hours of 

 rapid .nT.ition yielded 5.06 mg. of ("< ).j, or 0.23 mg. per hour. 



3. To learn whether some of the expired CO., was getting by 

 the collet ting jars, a barium hydroxide bottle as in the previous 

 ase was placed between the collecting jars and the pump. A 

 rapid Mream of air passing through two respiration jars and 

 t\\o seta <>l" i "lle< -ting jars (in parallel; one for Nectnrns, one for 

 Rdini />//>/V;;M. was sucked through this jar for 26 hours. 28.6 

 niv;. it ( ( ^ \\ere collected, or 1. 1 mg. per hour. Subtranmg 

 from thi> Injure the amount introduced into the jars with the 

 inhaled air. it appears that 0.87 mg. per hour of C() : was bein- 

 lost from the >\ >tem. The animals in the jars weighed 93 grain-. 

 so that the !<> i- 0.00936 mg. per hour per gram weight of the 

 animals. This i- about an 8 per cent, loss for Xectnrns ami 

 about 6 per cent, for the frog. For the larger animals an addi- 

 tional collecting jar was used. This of course tended to keep 

 down the loss. 



