316 APPLICATION Of METHODS OP BACTERIOLOGY 



is very small indeed; but for organisms that have already 

 been exposed for a time to such agents this amount is very 

 much less. It is plain, then, that if the test is to be an 

 accurate one, precautions must be taken against admitting 

 this minute trace of disinfectant to the medium with which 

 we are to determine whether the bacteria exposed to the 

 disinfectant were killed or not. The precautions hitherto 

 taken to prevent this accident have been, when the threads 

 were employed, washing them in sterilized distilled water 

 and then in alcohol; or, where fluid cultures were mixed 

 with the disinfectant in solution, an effort was usually 

 made to dilute the amount of disinfectant carried over, to 

 a point at which it lost its inhibiting power. 



While such precautions are sufficient in many cases, 

 they do not answer for all. Certain chemicals have the 

 property of combining so firmly with the threads upon 

 which the bacteria are located as to require other special 

 means of ridding the threads of them; and in solutions in 

 which proteid substances are present along with the bacteria 

 a similar union between them and the disinfectant may 

 likewise take place. In both instances this amount of disin- 

 fectant adhering to the threads or in combination with the 

 proteids must be eliminated, otherwise the results of the 

 test may be fallacious. A partial solution of the problem 

 is given through studies that have been made upon corrosive 

 sublimate in its various applications for disinfecting pur- 

 poses, and in this connection it has been shown by Shaefer' 

 that it is impossible to rid silk threads of the corrosive 

 sublimate adhering to them by simple washing, as the sub- 

 limate acts as a mordant and forms a firm union with the 



1 Berliner klin. Wochenschrift, 1890, No. 3, p. 50. 



