350 HARVEY M. SMITH. 



than normal. Berezowski (1910) worked on the size of the 

 intestinal cells of the white mouse during development, and 

 shows that these cells become larger as the animal grows. 



Krogh (1916) gives comparative tables on basal metabolism 

 as it has been worked out by various workers on different animals. 

 These figures indicate that there is a general agreement between 

 activity as measured by basal metabolism, and cell size as meas- 

 ured on the red blood cells by Gulliver (1875), in inverse ratio. 

 Figures given by different w'orkers on metabolism vary quite 

 widely, and this is true to an extreme degree of Amphibia. For 

 instance Regnault and Reiset (quoted in Morat and Doyon, 1900) 

 found that 0.063 m g- of CO 2 was eliminated per gram of frog 

 per hour, while Krogh (1916) gives a figure which corresponds to 

 0.3686 mg. of CC>2 per gram per hour. 



II. MATERIAL AND METHODS. 



As many different species of Amphibia were used as it was 

 possible to obtain. Activity was measured in terms of carbon 

 dioxide output. This was measured by fixing the gas as a pre- 

 cipitate of barium carbonate in a barium hydroxide solution. 

 In detail the method consisted in sucking air by means of a filter 

 pump through one 8-inch tube of concentrated potassium hydrox- 

 ide, two 8-inch tubes of soda lime and a gas washing bottle con- 

 taining strong barium hydroxide. This series was to remove the 

 carbon dioxide from the atmospheric air. The stream of air 

 then passed into a respiration chamber containing the animal. 

 Even when air breathing animals were used some water was 

 always placed in the bottom of this chamber to keep the skin of 

 the animal moist. From the respiration chamber the air current 

 passed through two or three gas-washing bottles containing a 

 carefully measured amount of standardized barium hydroxide. 

 Special care was taken to see that the air was broken up into fine 

 bubbles as it passed through these bottles. To accomplish this 

 the end of the inlet tube was drawn out into two fine points. 

 The bulb type of bubbler was found unstaisfactory because it 

 was too easily broken in the numerous manipulations incident to 

 making a long series of determinations; and because a finer 

 stream of bubbles could be obtained by the method described. 

 Suction tubing was used in making connections, and special 

 precautions were taken to avoid leaks. 



