172 BACTERIAL POISONS. 



must be added at this stage to prevent putrefaction, or the 

 dialysis can be carried on in a current of water which is 

 warmed to from 45° to 50° C, at which temperature the 

 growth of microorganisms is inhibited. After dialyzing 

 for twenty- four hours or more the greater part of the salt 

 will have vanished, and the alburaose will be found iu 

 solution in a considerable quantity of water which will not 

 have passed through the parchment. It is now necessary 

 merely to concentrate the solution and precipitate the 

 albumose by the addition of alcohol. In my earlier ex- 

 periments this was accomplished by evaporating in vacuo 

 at a temperature of 45° to 48°. When at length the liquid 

 has been reduced to a few cubic centimetres it is poured 

 into alcohol, and the precipitated albumose is filtered oif, 

 washed with the same reagent (alcohol), aiid dried. 



" Evaporating in vacuo is a long and tedious process, 

 and it requires a somewhat complicated apparatus. When 

 it is used for pathogenic albumoses there is always a risk 

 of the temperature employed destroying or diminishing 

 their physiological properties. Further, if the albumose 

 is allowed to evaporate to dryness, it may be difficult to 

 make it pass into solution again. To avoid these difficul- 

 ties I have designed a method of concentrating such solu- 

 tions which is less objectionable. It depends on the prin- 

 ciple that, if alcohol and water are placed on opposite sides 

 of a membrane, the water rapidly dialyzes through to mix 

 with the alcohol, while only traces of alcohol pass through 

 to mix with the water. Consequently, if a watery solu- 

 tion of albumoses is dialyzed against alcohol, the solution 

 diminishes in bulk and is rapidly concentrated, owing to 

 the passage of the water through the membrane. 



" My modus operandi is to place the dilute albumose 

 solution in a parchment sausage skin which is immersed in 

 a foot glass full of methylated spirit. The spirit can be 

 changed after some hours if it is desired to prolong the 

 process ; but this is not usually necessary. In this way I 

 have been able to bring 400 c.c. of albumose solution down 

 to 100 c.c. in the course of a single night, at the ordinary 

 temperature, without risk to the albumose or trouble to 



