54 ANNALS NEW YORK ACADEMY OF SCIENCES 



fresh water was found to be respectively 1.030 and 1.026. The specific 

 gravity determinations were made on fresh urine. One sample was large 

 enough for the use of the hydrometer. The other determinations were 

 made with the pycnometer. Denis ('12 ) records the normal urine of one 

 Mustelus canis as 1.032. 



Finally, the average amount of chlorine in two samples of normal urine 

 was 9.1812 gms. CI per 1000 c. c. urine, whereas the chlorine in the urine 

 of two other fishes immersed in a solution of one-half sea-water plus one- 

 half fresh water for about four hours amounted to 6.9517 gms. CI per 

 1000 c. c. 



It must be concluded, therefore, that the urine collected from fishes 

 immersed in diluted sea-water is more dilute than normal urine. What 

 is the concentration of the blood under these experimental conditions? 

 Are the activities of the kidneys such as to conserve in any way the os- 

 motic pressure of the blood? In the specimen whose urine had a A of 

 1.61° after about four hours' immersion in one-half sea-water plus one- 

 half fresh water, the A of the blood was 1.64°. The blood is slightly more 

 concentrated than the urine ; but it has already been shown that there is a 

 great reduction in salts in the urine of the fish immersed in fresh water. 

 This leads to the conclusion that the salts are not being excreted, but are 

 being held back by the excretory organ. The kidneys are acting to main- 

 tain the osmotic pressure of the blood by the excretion of water. The 

 problem is complicated by the fact that constantly water is coming into 

 and salts are leaving the blood through the gills. 



Presence of Salts in the External Medium after the Immersion 

 OF Fishes in Distilled Water 



If salts diffuse from the blood out through the gills, an analysis of the 

 diluted sea-water in which Mustelus is immersed should reveal the pres- 

 ence of these salts. To test this I made the following experiment : 



A male dog-fish 60 cm. long was pithed and a bolus of oiled cotton was 

 placed at the entrance of the stomach to prevent regurgitation of the 

 stomach contents. The fish was then immersed in sea-water. This was 

 gradually changed to fresh water in about five minutes, when the fish was 

 removed, thoroughly washed in fresh water and placed in a jar containing 

 two liters of distilled water. No urine was allowed to enter the jar. Air 

 was bubbled into this water during the course of the experiment. The 

 fish was near death when taken out of the jar forty-five minutes later. 

 The chlorine in one-half of this water was then determined in the follow- 

 ing manner : The sample was boiled down to 200 c. c. and filtered ; 20 c. c. 

 of the filtrate was analyzed for chlorides by the Volhard method. This 



