332 COLLOIDS IN BIOLOGY AND MEDICINE 



water and some of the crystalloids are absorbed, so that a concentrated 

 solution, the urine, passes off. 



In order to give an idea of the quantities of fluid which are in- 

 volved, the figures of H. MEYER and R. GOTTLIEB * are reproduced. 

 The blood contains about 0.6 per cent urea, the daily urine about 

 30 gm. There must thus be about 50 liters filtered through the 

 glomeruli and about 48.5 liters reabsorbed by the tubules. Since in 

 24 hours from 500 to 600 liters of blood flow through the kidneys, 10 per 

 cent of the water of the blood must be filtered off. This is not at 

 all improbable when we recollect that the afferent vessels (vas afferens) 

 have a much larger lumen than the efferent (vas deferens). 



The filtration processes are relatively the least difficult to explain. 

 The criticism until recently offered that no filtration could occur 

 through homogeneous colloid layers has been disposed of by the ultra- 

 filtration experiments of H. BECHHOLD.* 41 It must not be insisted 

 too strongly that the phenomenon in the glomeruli is a "filtration" 

 since it is evidently a process midway between filtration and diffusion. 



As in the case of every other ultrafiltration, that in the kidney is 

 dependent upon the pressure. According to E. H. STARLING, it 

 begins with an arterial pressure of at least 40 mm. mercury; below 

 this the secretion of urine ceases. In blood artificially diluted with 

 water, a minimal blood pressure which just maintains the circula- 

 tion suffices for the secretion of urine. 1 R. GOTTLIEB and R. MAGNUS 

 showed this by permitting normal saline to flow continuously into 

 an animal's vein. According to GOLL the urinary secretion rises 

 and falls almost proportionately to the blood pressure. The experi- 

 ment of D. R. HOOKER* furnishes a result in point. He found in 

 the isolated dog's kidney, that with a constant perfusion pressure 

 the quantity of urine formed was directly proportional to the size of 

 the (artificial) pulse pressure. The pulsation of the blood pressure 

 plays an important part in the rate of filtration, H. BECHHOLD * 12 

 compared the amount of fluid which flowed through an ultrafilter 

 under constant and under pulsating pressure, and found that in the 

 latter case the filtrate was considerably more than in the former. 

 We may imagine that with the lower pressure the filter absorbs the 

 fluid completely and it is pressed out again with the increased pres- 

 sure. This depends very much on the elasticity of the ultrafilter; if 

 it can follow the variations in pressure very rapidly, the rate of 

 filtration is higher than for an inelastic filter. Perhaps this will 



1 I cannot agree here with MARTIN H. FISCHER (Oedema, p. 209.) He says, 

 "Enormous pressures are necessary to filter fluids through thin colloidal mem- 

 branes." This is not correct. With suitably prepared thin collodion membranes, 

 a few centimeters of pressure suffice for ultrafiltration. 



