THE FORMATION OF LYMPH. 415 



and peptone in the presence of large quantities of other 

 proteids like albumins and globulins, such as occur in the 

 blood. The method consists in adding to the blood an 

 equal volume of a ten per cent, solution of trichloracetic 

 acid, and filtering. The precipitate consists of the albumins 

 and globulins, the filtrate contains the proteoses and pep- 

 tones. This method has yielded most valuable results in 

 the hands of subsequent observers, only, as C. J. Martin 1 

 points out, it is advisable to boil the mixture before filtering, 

 and filter hot, as in the cold proteoses are partly precipitated 

 by this reagent ; the precipitate, however, is soluble on 

 heating. 



Starling's next paper, written in conjunction with W. 

 M. Bayliss, relates to the subject of capillary pressure. 

 They point out that in Heidenhain's experiments on the 

 relation of lymph formation to blood pressure, the arterial 

 blood pressure is the only factor measured, and is spoken of 

 as if it were synonymous with or varied directly as the 

 capillary pressure. Now this can only be generally true if 

 the vascular system were composed of tubes through which 

 fluid was flowing from an inexhaustible reservoir. But the 

 vascular system is a closed system of tubes with definite 

 capacity, and containing a definite amount of fluid, and 

 between the capillaries and the arterial pressure is situated 

 the peripheral resistance, that is the arterioles, so that unless 

 we are certain that these arterioles do not vary in calibre, it 

 is impossible to say that a rise of arterial pressure will 

 occasion a corresponding rise in the capillaries beyond the 

 peripheral resistance. The venous pressure would be a 

 better measure of capillary pressure, but it is best of all to 

 take simultaneously arterial and venous pressure, that is the 

 pressure on each side of the capillaries, since it is not 

 practicable to measure the capillary pressure directly. 

 Remembering this principle, but without going into the 

 details of the experiments, it will be useful next to glance 

 through the following table, which gives the main con- 

 clusions arrived at : — 



^■Journal of Physiology, xv., 375. 

 29 



