viii THE EXCRETION OF URINE 439 



by reabsorptiou of water, so that the pigment is precipitated and 

 penetrates the striated cells of the convoluted tubules. In fact, 

 Sobieranski observed that in these cells the stain is most con- 

 spicuous near the inner surface, as though the pigment had 

 entered by the lumen of the tubules and not by the lymph 

 channels. 



This view is confirmed, according to Sobieranski, by the effects 

 of injection with carmine, which, as it is less diffusible than the 

 sulphindigotate of soda, can more easily be followed in its passage 

 through the kidney. If the rabbit be killed 30-40 minutes 

 after intravenous injection of carmine, and the pigment fixed 

 by injection of absolute alcohol through the renal vessels, the 

 glomeruli stain red, and the epithelia of the convoluted tubules 

 are found to contain pigment granules in the part facing the 

 lumen, never in the basement membrane facing the lymph spaces. 

 According to Sobieranski, this shows that the carmine is taken up 

 by the striated cells on the side of the lumen, with the water they 

 absorb. But even if these facts minimise the value of Heiden- 

 hain's argument, they do not, in our opinion, prove as much as 

 Sobieranski claims, when he attempts to reinstate Ludwig's theory 

 by saying that the tubular epithelium is an apparatus for con- 

 centrating the glomerular filtrate since this concentration is 

 due not to osmotic processes, but to a special physiological activity 

 of the cells, analogous to that of the intestinal epithelium as 

 expressed in an internal secretion. 



In fact, we may suppose that the pigments injected into the 

 blood (both sodium sulphindigotate and carmine) are specially 

 eliminated by the epithelial cells of the uriniferous tubules, 

 and, when present in excess, by the glomeruli also; and that 

 the internal part of these cells stains more than the basal part 

 may depend on the fact that the cells habitually expel the whole 

 of the substances which they take up from the lymph into the 

 duct as fast as they absorb them. Immediately after the death of 

 the animal, therefore, the whole of the substances already absorbed, 

 or able to be absorbed before the extinction of vital activity, are 

 collected in the inner portion of the cell, and are partly excreted 

 into the lumen. 



In order to decide which of the two theories holds good, that 

 of Bowman and Heidenhain, who allow a physiological function 

 of external secretion to the epithelial cells of the tubules, or that 

 of Ludwig as modified by Sobieranski, who attributes to these cells 

 a physiological function of internal secretion, it was necessary to 

 discover some method for studying the function of the uriniferous 

 tubules apart from that of the glomeruli. Nussbaum (1878) 

 endeavoured to solve this problem by certain experiments on the 

 frog which may be summarised as follows : 



The amphibian kidney possesses a double series of vessels that 



