PERFORMANCE OF THE EXPERIMENT 183 



with too much toluol, or if the tube at the beginning 

 of the test has not been immersed sufficiently deeply, 

 then it may easily happen that, during the intro- 

 duction of the pipette, some of the liquid passes from 

 the outside into the dialysing tube. While, if one 

 sucks strongly with the pipette at that moment, then 

 the reverse may occur, and the contents of the tube 

 may enter the pipette. 



Ten cubic centimetres of the dialysate are taken, by 

 means of a closed pipette passed through the toluol 

 layer, and poured into a dry, wide, and absolutely 

 clean test-tube. For each dialysate, as a matter of 

 course, a separate, absolutely clean and dry pipette is 

 used. One's work should never be arranged in such 

 a manner that the pipette has to be hurriedly cleansed 

 with alcohol, water, or ether after use, for the clean- 

 ing in this case may very easily be insufficient. 



The danger of soiling the pipette with saliva is 

 particularly great. (Compare p. 153.) 



Then o'2 c.c. of the i per cent, aqueous solution of 

 ninhydrin are added to each test, together with a dry 

 boiling rod (see p. 159), and one test after another is 

 boiled absolutely evenly for a whole minute (see 

 p. 1 60). After half an hour we ascertain which tests 

 show a coloration and which do not, and only then do 

 we compare our results with the original dialysates. 

 If there be any tests which have evaporated more than 

 the others, they are rejected, if they show a positive 



