FILTRATION OF BACTERIAL LIQUIDS. 



469 



vary somewhat in thickness, and hence the rate of flow is 

 variable. Highly albuminous fluids cannot be filtered un- 

 less the wall of the filter is relatively thin or unless ex- 

 tremely high pressure is employed to force the soluble pro- 

 teins through the pores of the filter. With insufficient 

 pressure these remain on the filter and 

 only an aqueous liquid will pass through. 



The liquid to be filtered should con- 

 tain as little suspended matter as possible. 

 For this reason, blood-serum should be 

 eentrifugated or allowed to settle until 

 free from corpuscles. Bouillon cultures 

 should first be filtered through several 

 thicknesses of paper.' The hardened, 

 parchment-like paper (No. 575) of Schlei- 

 cher and Schiill is well adapted for remov- 

 ing the mass of bacteria from a liquid. 



A large number 

 of filters have been 

 devised, but most of 

 these are far from 

 being satisfactory 

 for ordinary labora- 

 tory work. One of 

 the best 'filters is that of Martin, but it can be used only 

 with negative pressure, — that is to say by producing a 

 vacuum in the receiving fiask. The author has described 

 a filter' which has been found to be very useful for labor- 

 atory work. It can be connected with an air pump, or 

 additional pressure may be obtained by connecting with a 

 cylinder of compressed air. It can be used for filtering 

 large or small quantities of liquid. 



Fig. 66. Apparatus for filtering bacterial liquids {F. G. N.). 



The apparatus shown in Fig. 66 consists essentially of a glass 

 cylinder (g) 3 cm. in diameter and 20 cm. in length. The upper end is 



1 Centralblatt fiir Bakteriologie 22, p. 337, 1897. 



