570 Proceedings of the Royal Society 
7 per cent., for by calculation 1 gramme of urea should yield 300 c.c. 
of nitrogen at 0° C. and with 760 mm. of pressure. 
It only yields, however, in practice, 340 c.c., the deficit is, however, 
very constant. 
By adopting a modification of this, for which I am indebted to 
Professor A. Gamgee, almost all the nitrogen is given off. This 
is the addition of ordinary cane sugar syrup, when you obtain 
363*4 c.c. of nitrogen. 
This process was, of course, tested with great care in order to find 
out whether it was sufficiently accurate. Two equal portions of 
the same blood were taken, and to one was added a known 
quantity of pure dry urea. If the method was perfectly ac- 
curate the results of the two fluids when analysed should show a 
difference equal to the quantity of urea added. 
That this was not the case is evident ; for although the method 
approaches perfection, yet it is far from being perfect, and it 
would only lead to disappointment if its faults were passed lightly 
over. 
Solutions of urea have still to be evaporated although in a com- 
paratively pure form and with far less fluid. This, as we have 
shown, means loss. 
This loss, however, is brought to a minimum, and is as a result of 
my test analyses not more than 7 per cent, or 8 per cent. Now, as 
part of this loss is constant, and as urea is actually obtained and 
estimated as such, the fallacy is less than these numbers would 
indicate. 
That this method will ere long be supplanted by another and a 
better one I have no doubt, and, indeed, while working at its appli- 
cation to the investigation of physiological conditions, I am at the 
same time continually trying to improve the method. So exact, how- 
ever, is it that one may always readily obtain a demonstration of 
urea from so small a quantity of blood as 10 c.c. I have here a 
specimen of urea in a very pure form from 15 c.c. : this has been 
frequently purified, and is mixed now with little else ; one might 
almost estimate its quantity by weight. I may state that in blood 
far more urea is present than is ordinarily imagined. I have 
obtained 56 parts per 100,000, which is a very large quantity. The 
normal is between 40 and 30 part 100,000. 
