determining the amount of Urea in the Urinary Secretion. 389 



evolution of gas is produced ; and the quantity of hypochlorite 

 used to arrive at this point indicates the amount necessary. I 

 found by experiment that one grain of urea requires somewhere 

 about half a fluid ounce of the ordinary sodse chlorinatse liquor 

 for its complete decomposition. The amount of mercury era- 

 ployed requires some little attention. It should, as a general 

 rule, be never less than the volume of gas produced ; for if the 

 volume of gas evolved is more than that of the mercury used, it 

 will be more than that of the solution of salt, and therefore some 

 of the mixture of urine and hypochlorite will be forced out of 

 the tube before it is completely decomposed, and consequently 

 some of the gas will be lost ; so that if this occurs, we must 

 repeat the experiment, using either a larger quantity of mercury 

 if our tube will allow, or diminishing the quantity of urine em- 

 ployed. 



It might be supposed on first sight that this method would be 

 liable to the following source of error, viz. that some of the gas 

 would be evolved and lost during the pouring in of the hypo- 

 chlorite ; but this is not the case, as several seconds elapse before 

 there is any apparent reaction or evolution of gas on mixing the 

 hypochlorite with the urine, and there is therefore full time to 

 perform the experiment without any loss of the gas. I have 

 also ascertained that the acid reaction of the urine does not 

 affect my method. I should observe that this new method, like 

 all the others known, is not perfectly free from some slight 

 sources of error ; the principal one being, that ammonia, if it 

 exists in the urine, gives rise to nitrogen gas, and therefore 

 increases the apparent amount of urea ; but the same objection 

 holds equally in Liebig's and Ragsky's methods, which are per- 

 haps the two most accurate at present known. Uric acid also is 

 similarly affected by the hypochlorite ; but it and ammonia ordi- 

 narily occur in such small proportion in urine, that the error 

 produced from these substances would be but trifling, and is 

 partly corrected by taking the calculated quantity of nitrogen, 

 which is, as I have shown, something more than that obtained 

 from a certain quantity of urea by direct experiment. 



In cases where ammonia or uric acid occurs in more than ordi- 

 nary quantity, these substances must be separated by the usual 

 means employed before having recourse to my method. 



1 should think that gently heating the urine with a certain 

 quantity of baryta water as long as the odour of ammonia is 

 disengaged, and then filtering the solution, as recommended by 

 Liebig, for the separation of aumionia before applying his method 

 (see the Quarterly Journal of the Chemical Society, vol. vi. p. 30), 

 would effect the object very easily, and separate not only the am- 

 monia, but also the greater part, if not all of the uric acid present. 



