LYMPH. 367 



sage of crystalloid bodies from a less concentrated to a more concentrated solu- 

 tion. Cohustein 1 has endeavored to show a fallacy in these results. He con- 

 tends that since it requires some time (several minutes) for the lymph to form 

 and pass into the thoracic duct, it is not justifiable to compare the quantitative 

 composition of specimens of blood and lymph taken at the same time. If one 

 compares, in any given experiment, the maximal percentage in the blood of 

 the substance injected with its maximal percentage in the lymph, the latter 

 will be found to be lower. This, however, does not seem to be the case in all 

 the experiments reported. The work of Mendel 2 with sodium iodide seems to 

 establish the fact that when this salt is injected slowly its maximal percentage 

 in the lymph may exceed that in the blood ; and in the experiments made by 

 Cohnstein, as well as those by Mendel, it is shown that the percentage of the 

 substance in the lymph remains above that in the blood throughout most of 

 the experiment. In this point, therefore, there seems to be a real difficulty in 

 the direct application of the laws of filtration and diffusion to the explanation 

 of the composition of lymph. It is possible, however, that a better under- 

 standing of the conditions prevailing in the capillaries with regard to osmosis 

 and filtration may clear up this difficulty. 3 Meanwhile it seems evident that 

 in spite of the very valuable work of Heidenhain, which has added so much 

 to our knowledge of the conditions influencing the formation of lymph, the 

 existence of a definite secretory activity of the endothelial cells of the capil- 

 laries has not been proved. 



1 Archivfilr die gesammte Physiologic, 1894-95, Bde. lix., Ix. und Ixii. 



2 Journal of Physiology, 1896, vol. xix. p. 227. 



3 See Hamberger: Du Bois-Beymond? s Archivfiir Physiologic, 1896, S. 36. 



