CELL SIZE AND .METABOLIC ACTIVITY IN AMPHIBIA. 357 



from starvation. Xo. 18 was starved, while No. 19 was used 

 for blood smears within a few days after being collected. 



Ilyla picker ingii (Holbrook). One specimen was collected and 

 blood smears made before the carbon dioxide tests had been 

 started. None were available while tests on metabolism were 

 being made. 



Acris gryllus (Le Conte). Two specimens were collected from 

 the field and blood smears made. Time was lacking to make 

 carbon dioxide output tests. 



Chort>philns iiigritus (Le Conte). See note on Ilyla pickeringii. 



/in 1 " iitnericanus (Le Conte). Two specimens were collected 

 .ind ~t. nved at room temperature three months before using. 

 No. 3 i He. | ot -tarvation after only one test for carbon dioxide 

 h.nl U Tii made, and blood could not be obtained for smears. 



IV. CONTROLS. 

 .1 . Cell Size. 



i. Blood corpuscles may shrink as smear- dry. However the 

 -ame methods were used on all specimens, so that the results 

 -liould be comparable. That there is a certain amount of differ- 

 ential shrinking due- to differences in thickness of smear is suv;- 

 i;c>icd by comparing measurements from different slides made 

 from the same individual. For instance the five slides of blood 

 Irom Kaiia clamitans No. 2 gave the following series of averages 

 of 50 ( oi pii-cles from each slide: 23.98x13.02, 22.92x11.66, 

 ;j \ u.o.s.23.1 \ 12.32 and 22.86 x 12.02. Some of this varia- 

 tion may have been caused by differences in the corpuscles which 

 happened to In- measured, but most of it was probably due to the 

 nature of tlif -mear. With the slide which averaged 23.98 x i.V<>- 

 tlu- tir-t J5 i i-i-|)uscles measured 23.84 x 13.12, while the la-t 25 

 measured 24-U\ u.()2. The method used is that ot" (iulli\cr 

 i ^751 except for tlir tact that he does not appear to h.i\e Mained 

 his -meais. < .. -i ^. 'polus (1906) States that dry preparations 

 are unrelialile liecau.-e the sixe of the cells is likely to vary with 

 tile thirkne-- ot the film, and states his preterem c tor tin- \\et 

 method. Thi- icn-i-ts in placing a small drop of fresh blood 

 on a clean >lide and quickly placing on a cover. The corpu>cles 

 are then measured immediately. Thi> method has been found 

 u-eles- for Amphibia because the corpuscles are distorted by the 



