OXYGEN CONSUMPTION OF KIDNEY 159 



The saline diuresis may have been caused by all or any of these 

 concomitants. They may be eliminated one by one. 



(1, 2, and 3) Increased blood volume, pressure and flow may 

 be considered together. Increase in pressure, etc., produced 

 mechanically without altering the concentration of corpuscles or of 

 colloids, certainly does produce an increased flow of urine, the 

 constituents of which have a concentration approximating to 

 that produced after injection of Ringer's solution. Barcroft and 

 Straube overcame this difficulty very ingeniously. 



They previously removed a quantity of blood equal in volume 

 to the Ringer they were about to inject, thus keeping the blood 

 volume, etc., normal. The diuresis was produced as before, 

 entailing no extra oxygen consumption. 



(4) The addition of blood corpuscles to make up the deficient 

 concentration made no appreciable difference in the flow of 

 urine. 



(5) Knowlton introduced a colloid, gum acacia or gelatine into 

 the perfusion fluid so that the colloidal osmotic pressure of the 

 injected fluid was equal to that of the blood (25-30 mm. of Hg.). 

 This prevented the onset of marked diuresis, gelatine being more 

 efficient in this respect than gum acacia. Two causes may be 

 ascribed to the lower efficiency of gum : (a) its lower osmotic 

 pressure (5 per cent, gum has an osmotic pressure of about 12 mm. 

 compared with 5 per cent, gelatine whose O.P. is 23 mm. of Hg. 

 Bayliss recommends a 7 per cent, solution), (b) Gelatine has a 

 certain water holding power which is altered by treatment with 

 salts (see Imbibition, Chap. VIII.). 



The conclusion that one would draw from this series of experi- 

 ments is that the passage of fluid and salts through the kidney by 

 filtration is controlled by the concentration of colloids in the 

 blood plasma. 



Let us consider now a case where oxygen is consumed and, 

 presumably, work done. Sodium sulphate is a diuretic. It is 

 less diffusible than sodium chloride. A collodion membrane can 

 be prepared which will allow the chloride and not the sulphate to 

 pass through. Yet in the kidney the very reverse seems to take 

 place. Sodium chloride' acts very much like Ringer's solution, 

 and though after a large injection of NaCl the chloride content 

 of the urine rises, it does not materially alter the concentration 

 of the other solutes. On the other hand, the injection of Na 2 SO 4 

 is followed by the secretion of a urine almost entirely a water 

 solution of the sulphate. Cushny has shown that, after the 



