SCOTT, STUDY OF CHANGES IN MUSTELVS CANIS 49 



and assistance in the chemical technique here involved. The chlorine in 

 the blood is an index of the salts present. As a check on the method the 

 chlorine content of successive samples of blood from five fishes taken from 

 sea-water was determined. For purposes of comparison the quantity of 

 chlorine present is expressed in grams per 1000 grams of blood. The 

 average amount of chlorine in the first sample of blood taken from each 

 of the five specimens was 6.597 grms. p. m. The average amount of 

 chlorine in the second sample was 6.668 grms. p. m. The difference is 

 within the limits of experimental error. The analysis corroborates the 

 results obtained by measuring the freezing points of successive samples 

 of the blood of the dog-fish taken from sea-water. In a second series of 

 experiments, after a normal sample of blood had been taken from each of 

 five fishes, the fishes were placed in a concentrated solution of sea-water 

 having a A of about 3.15° for one hour, at the end of which a second 

 sample of blood was drawn from each specimen. The average amount of 

 chlorine from the normal blood amounted to 6.249 grams per 1000 grams 

 of blood. The average amount of chlorine in the blood after the immer- 

 sion of the fishes in the concentrated sea-water was 7.522 grams p. m. 

 A gain of 20.4 per cent in chlorine is indicated, which under these condi- 

 tions probably means a gain of 20.4 per cent in salts. In the third place, 

 an analysis was made of the chlorine content of the normal blood of 

 twenty specimens of Mustelus canis. In some cases the blood of two or 

 even three specimens was mixed for a single analysis. Analyses were also 

 made of the blood of twenty fishes after immersion in fresh water. The 

 average quantity of chlorine in the normal blood was 6.098 grams p. m., 

 while the average quantity of chlorine in the blood after immersion of 

 the animal in fresh water for about one hour was 4.638 grams p. m. This 

 means that the blood had lost 23.9 per cent in chlorine; but in this case 

 also the loss in chlorine probably means an equivalent loss in salts. It has 

 thus been shown that on immersion of the animal in a concentrated solu- 

 tion of sea-water the blood gains in chlorine ; on immersion in fresh water 

 the blood loses in chlorine. 



In order to avoid possible errors due to the volatilization of chlorides 

 through ashing, I decided to make an analysis of the serum of the dog- 

 fish under the above described experimental conditions. 



After the blood was drawn in each case, it was first defibrinated and 

 then centrifuged. The supernatant serum was drawn off with a volu- 

 metric pipette. In some cases it was necessary to use the mixed sera of 

 two specimens for an analysis. Five c. c. of serum was placed in a volu- 

 metric flask of the capacity of 100 c. c. About three c. c. of pure acid was 

 added. The flask was half filled with distilled water and the contents 



