METHODS OF COLLOID 



L is the length of the capillaries (i.e., not smaller than the thick- 

 ness of the wet filter). 



k is a constant factor dependent on temperature and kind of fluid. 

 The following formula applies: 



Q (R + /) L 



D 



k-S-F-R 



If all the experiments are performed under the same conditions, 

 the formula may be simplified, because , g becomes a constant. 



K * O V 



For the practical performance of this experiment two persons are 

 required, one of whom regulates the pressure, while the other de- 

 termines the amount of water filtered over a certain time, fixed by 

 means of a stop watch. 



Under the apparatus is placed a funnel which has a rubber tube 

 with a pinch-cock attached. 



The ultrafiltration apparatus is filled with water, and the air pres- 

 sure is raised to a given point. The pinch-cock is then closed so 

 that all the water filtering at a constant pressure is caught in the 

 funnel. After a given time (e.g., one minute) has elapsed, the entire 

 pressure is instantly released. In this way the amount of water 

 filtered through a given filter in a given time is measured. If we 

 have previously performed the same experiment with a filter paper 

 which has pores of known size, one which, for instance, even par- 

 tially retains blood corpuscles or bacteria, i.e., objects which are 

 measurable microscopically, by means of the above formula, we 

 can estimate the average size of the pores of the ultrafilter. 



With this method ultrafilters which just held back hemoglobin 

 showed the average diameter of their pores to be from 33 to 36 JJLJJI. 



4. Method of Emulsion Filtration, described on pp. 15 and 16. 



Adsorption by Filters. In ultrafiltration experiments, it is 

 necessary to avoid errors due to adsorption on the part of the filter. 

 Accordingly, as a preliminary experiment, it is advisable to shake a 

 portion of the solution with a shredded filter. If the content of 

 the solution is practically the same -afrte "as before the shak- 

 ing, there has been no adsorption. If the adsorption introduces an 

 error into the ultrafiltration experiment, it is necessary to use a dif- 

 ferent jelly. For instance, arachnolysin is very strongly adsorbed 

 by glacial acetic acid-collodion, but only slightly by formol- 

 gelatin. 



In ultrafiltration experiments it is always important to work 

 quantitatively and to test what remains in the filter as well as the 

 filtrate obtained. 



