Fir.TRATION OF IMMUNE SERUMS. 



339 



respectivelj'', were washed b}^ passing distilled water through them and 

 were sterilized in the autoclave. 



Three hundred cubic centimeters of previously filtered serum inocu- 

 lated with these small bacilli were passed through each of the nine sterile 

 filters under various pressures ranging from one to four atmospheres; 

 the portions of the serum were then collected in a sterile receptacle 

 provided with a protected mouthpiece and stopcock at the bottom, and 

 then bottled in two portions, a and h, of about 150 cubic centimeters 

 each. Portion a was drawn off into a sterile bottle as soon as about 150 

 cubic centimeters had collected, and portion & was bottled after all serum 

 including the froth obtained in the end had passed through the filter. 



After two days' incubation 1 cubic centimeter of each portion of the 

 filtrate was planted in liquefied agar. 



Table V. 



V, 



Vn 



Vim 



N, 



N„ 



N„, 



Wi 



W„ 



w,„ 



Filter. 



Portion. 



ja 

 lb 



K 



i: 



Pressure. 



1 atmos- ! 2 atmos- 

 phere. ! pheres. 



3 atmos- 

 pheres. 



4 atmos- 1 

 pheres. i 



-T- 



+ 



+ 



+ 



■V 



■f 



The foregoing Table Y shows the results that were obtained. All 

 filtrates from Berkefeld filter V yielded growths of these small bacilli, 

 they frequently passed through filter jST, while filter W arrested them 

 under one and under two atmospheres pressure, except in portion h from 

 filter III under two atmospheres pressure. Throughout the erperiment 

 contaminations were more frequent in portion h than in jiortion a. 



CONCLUSIONS. 



Altliougli the foregoing tabulated records of experiments are very 

 limited in extent, I feel justified in saying that with proper care immune 

 seriims can be passed through germ-proof filters without suffering any 



