Since the original publications of Liebig on the use of a solution of mercuric nitrate in the urea estimation several attempts have been made to so modify the process as to render it more perfectly accurate for the purpose. Liebig himself recognized the variable character of the precipitate formed when a solution of mercuric nitrate is added to one of urea, even when the latter has a mean concentration of 2 gm. to 100 c.c. and the mercuric solution is dilute, with 1 c.c. about equivalent to 10 mg. of urea.
This irregularity depends in part on the liberation of free nitric acid in the operation, and Rautenberg and others1 attempted to correct for this by the gradual addition of alkali in the form of calcium carbonate. In this paper the authors recommend to correct the direct result of the titration by subtracting 0.06 c.c. of mercury solution for