PERFORMANCE OF THE EXPERIMENT iSl 



i 



owing to the dilution of serum so caused. The tissue 

 should never be handled with the fingers. 



To the tubes containing the tissue i to 1*5 c.c. of 

 serum are now added. A rule should always be made 



j 



of arranging this experiment first. Afterwards from 

 i to i '5 c.c. of serum are placed in an empty tube 

 (control test). Then the tubes are thoroughly rinsed 

 with distilled water (as described on p. 151), and 

 placed in Erlenmeyer flasks which have previously 

 been filled with about 20 c.c. of sterilized water. Then 

 a large amount of toluol is poured into the tubes and 

 over the liquid outside, care being taken that the 

 part of the tube which projects from the liquid should 

 be soaked with toluol. At this stage of the experi- 

 ment the following sources of error may arise. First 

 of all, water may get into the tubes while they are 

 being rinsed. If the work is not carried on in a 

 scrupulously accurate manner considerable dilutions 

 may occur. The tube must be completely closed 

 during this operation. I have lately been in a 

 position to observe a second source of error which 

 may arise. Contrary to instructions the flask was 

 filled with 20 c.c. of water and a large quantity of 

 toluol, and only then was the tube and its contents 

 immersed. In this case the liquid in the flask was 

 raised to such a level that it passed from the outside 

 to the inside of the tubes. Besides, the tube came 

 into contact with the neck of the Erlenmever flask 



