closed at its upper extremity by the finger and rapidly passed through the layer of toluol, 10 c.c. of the dialysate are taken out, and placed in a test-tube bearing the same number as the corresponding Erlenmever flask. This is the best way to avoid mistakes. Of course, for each dialysate a separate and absolutely clean pipette must be used. We do not recommend transferring the dialysates to the test-tubes by means of the same pipette, rapidly cleansed each time after use, because by this means some impurity or other may easily be introduced into the dialysate. Some saliva may very easily enter that part of the pipette which, during the so-called cleansing, remains untouched by the water, alcohol and ether. On the contrary, new saliva is drawn in at each operation if the suction is made by the mouth. Now, when the dialysate is taken up, it is almost certain to be drawn above the level marked upon the pipette, and may then become mixed with the saliva. If test-tubes, graduated to 10 c.c., are to hand, then these tubes may be employed in the following manner: After removing the dialysing tubes, the toluol is drawn off and the dialysate is poured directly into the test-tube. It is of no great importance, in the biuret reaction, to consider quantities to the minutest exactness, nor does a little toluol do any harm.
Now, to each test-tube is added about 2.5 c.c. of a 33 per cent. caustic soda solution. The whole is shaken
