STERILISATION. 1 7 



of the author has shown. It is certainly not allowable to lay 

 it down as a general rule that beer must always be spoilt by 

 filtration. 



The filtration of milk of any biological importance has 

 proved, so far, impossible, as a fitter with pores large enough 

 to allow the fat globules to pass will not retain bacteria, of 

 which the vast majority are smaller than the globules. The 

 filter is, therefore, only of use in removing the greater part 

 of the dirt particles from the milk, and the micro-organisms 

 that are attached to them. On a large scale sand and gravel 

 filters are used. For instance, in the Danish system, as 

 constructed by Busck, a vertical cylinder is used with per- 

 forated diaphragms, between which are packed layers of 

 sand, the grains of which are coarser at the bottom and finer 

 at the top. The milk is run in from below. In Krohnke's 

 construction the milk is passed through a cylinder partially 

 filled with gravel, and carrying vertical diaphragms ; the 

 cylinder is rotated round a horizontal axis. On a small scale, 

 the fresh warm milk may be filtered through cotton wool, 

 a layer of which is inserted between two sieves ; the filter 

 requires renewing daily. A more complete biological puri- 

 fication has been attempted by pasteurisation. 



The filtration of air is intended, not only to remove living 

 germs, but also to remove all floating particles. It has already 

 been stated that Schroder and Dusch accomplished this by 

 means of a tube filled with closely packed cotton wool, and 

 this still proves to be one of the best materials. In the labor- 

 atory such filters are used to seal test tubes and flasks. When 

 they take the form of glass tubes, as in the Freudenreich 

 flasks, it is unnecessary to protect the surface at the open 

 end of the tube, but it is otherwise in the case of test tubes, 

 where a great part of the filter is exposed to the dust of the 

 air. Germs may easily grow on the cotton wool when it 

 absorbs moisture. Flaming the surface is not always sufficient, 

 and in such a case it is desirable to keep the tubes in an atmo- 

 sphere free from germs. By the diffusion of air, which goes 

 on through such small filters, evaporation takes place, and as 

 a consequence the liquid becomes more concentrated, or the 

 gelatine hardens on the surface. Such evaporation can be 



2 



