1912] On the Distribution of Potassium in Renal Cells 397 



raise or to lower surface tension at the liquid-air surface. As far as has 

 been investigated the tendency is for all substances in solution under 

 these conditions to concentrate at the interface where the tension is 

 lowered. 



The degree of concentration has been investigated experimentally 

 by Lewis^ in the case of a number of solutes. For this purpose he 

 allowed a known volume of a hydrocarbon oil to ascend through 

 aqueous solutions in a tall cylinder. As the diameter and, consequently, 

 the number of the droplets were ascertained the total surface area 

 they presented was also known. The oil was collected at the top of 

 the cylinder in a suitable apparatus. Each aqueous solution contained 

 a substance soluble in water but insoluble in oil, viz., sodium glyco- 

 cholate, methyl orange or Congo red. These lowered the surface 

 tension in proportion to their concentration in solution. The degree 

 of adsorption of each of these solutes on the surface of the oi' droplets 

 was from twenty to eighty times the theoretical value calculated 

 according to the Gibbs formula. Lewis further investigated the effect 

 of electrolytes^ and found they manifested a similai but less m?rked 

 condensation, the adsorption of potassium in potassium chloride 

 solution being thirty times that calculated. Indeed, in all cases the 

 adsorption was such that a process of gelatinisation at the surface 

 was suggested as an explanation. It would further appear that all 

 substances in solution are adsorbed at a fluid-fluid or fluid-solid 

 interface. 



Mayer^ has shown that the surface tension of blood plasma is lower 

 than that of an isotonic salt solution, and this would favour adsorption 

 on cell-lymph interfaces. We may, therefore, perhaps be justified in 

 concluding from this that the concentrations of potassium on the renal 

 tubules, which we have described, are the result of surface tension differ- 

 ences on the surface of the tubules. In the production of these differ- 

 ences metabolism may play a part. We know that the oxygen intake 

 in the kidney is very high, indicating a considerable liberation of energy, 

 which the cells may transfoim to surface energy, using it either to obtain 

 material for metabolic purposes or in the functional activities of excre- 

 tion. The excess absorption of oxygen during the period of diuresis 

 over the period of rest in the experiments of Barcroft and Brodie^° is 

 equal to 0.0401 c.c. O2 per gram per minute. The secreting area for dog's 

 kidney has not been calculated, but for comparison the area of a human 

 kidney of similar weight may be given. According to Piitter^^ the 

 secreting area of a kidney of 45 gr. is 2.2 m^ Transforming the excess 

 of oxygen consumption into dynes per cm'^., after allowing 1/20 for 

 concentration of urine, we find an expenditure of 271.3 dynes per cm^ 



