PHILIP H. MITCHELL AND J. WALTER WILSON 



49 



cal error and are due to something more than a mere seasonal variation 

 since ina group of twelve analyses on eight summer frogs the maximum 

 variation was 17 per cent of the average potassium content of the 

 dried muscle while the analytical error as indicated by duplicate 

 analyses of the same muscle or by analyses of the two fresh gastroc- 

 nemii, of the same frog did not exceed a maximum of 2.3 per cent 

 and averaged 1.37 per cent. 



The effect of perfusion with a potassium-free Ringer solution 

 on the potassium content of muscle was contrasted with that of 

 similar perfusion with Ringer solution. In each case both legs were 

 perfused during 15 minutes with about 300 cc. of solution until the 

 venous outflow appeared bloodless ; the artery of one leg was then tied 



TABLE II. 



The Potassium Content of Perfused Muscles. 

 Results Are in Per Cent of Potassium in the Moist Muscle, as Weighed. 



off and 1500 cc. of solution perfused in If hours through the other leg. 

 Results are given in Table II. 



The potassium content of the moist muscle was decreased in both 

 experiments. Fahr (8) has shown that frog muscle immersed in 

 Ringer solution during 20 hours contained a smaller percentage of 

 potassium than fresh muscle, though after a similar immersion in 

 isotonic sugar solution a greater potassium loss was shown. The 

 lower per cent of potassium obtained after perfusion seemed attribut- 

 able in part, at least, to the higher water content. This possibility 

 was indicated by the obvious distention and edematous appearance 

 of the much perfused leg. Edema as a result of perfusion with a 

 number of physiological salt solutions has been observed by Guntz- 

 berg (9), Abel (10), and others. Another experiment was therefore 



