98 On tJie Estimation of Uric Acid in Urine. [June 1 6, 



upon the use of a purer ammonium chloride, but chiefly upon a 

 precaution in Ludwig's method to which reference is made below. 



Notes on other Processes. 



Having regard to the discussion at present going on with respect 

 to the whole question of uric acid estimation, it may not be out of 

 place to chronicle such experience of other processes as has been 

 gained in the course of working out the method just described. 



In comparing the ammonium chloride figures with those of Lud- 

 w ig' s process, the experiments marked Series 1 above indicated an 

 excess by the former method equal to some 3 milligrammes from the 

 100 c.c. of urine. As stated above, a part of this excess was due to 

 impurities in the ammonium chloride, but it was mainly due, I 

 believe, to a destruction of nric acid which occurs in Ludwig's pro- 

 cess when the silver urate is treated with the alkaline sulph-hydrate. 

 It is, of course, well known that uric acid is decomposed on heating 

 with caustic alkalies, but it is hardly recognised with what great 

 ease the decomposition takes place. I find that, on boiling 50 milli- 

 grammes of pure uric acid briskly for five minutes with 10 c.c. of 

 Ludwig's sulphide solution (vide Neubauer and Vogel, * Analyse des 

 Harns,' p. 543), i.e., the amount used to decompose the silver salfc 

 from 100 c.c. of urine, there is always an appreciable loss, sometimes 

 amounting to 3 milligrammes. It is therefore obvious that even five 

 minutes' boiling would be fatal to the accuracy of an estimation. In 

 Series 2 above, the silver compound was thoroughly shaken up 

 with the sulphide in the cold, and only just heated to boiling before 

 filtration. 



Mr. E. Groves, B.Sc., has recently described ('Journal of Physio- 

 logy/ December, 1891) a modification of Ludwig's method which I 

 have several times employed. Mi'. Groves uses potassium iodide 

 instead of the sulphide to decompose the silver urate. In my hands 

 the method has always given results which are decidedly too low, 

 owing doubtless to the action of the liberated iodine upon the uric 

 acid in the acidified filtrate. 



I have performed a large number of experiments with Haycraft's 

 method, and have found that results obtained by its means are 

 invariably too high, while the proportionate error undoubtedly 

 appears to be variable. This appears to be the common experience. 

 My own experiments, however, lend no support to the extraordi- 

 narily wide variations from the truth which some observers, espe- 

 cially in England, have ascribed to it. I should put the limit of 

 error at from + 10 to + 15 per cent. 



In conclusion, it may be claimed for the separation by means of 

 saturation with chloride of ammonium that it yields results which 



