ADVANCED PHYSIOLOGICAL CHEMISTRY 467 



white precipitate adhering to the sides of the beaker. On microscopic 

 examination the crystals consist of fine needles, which are grouped 

 into rosettes or sheaths (see Fig. 273). The precipitate is collected, 

 either on a weighed filter (see p. 492), or, if it be desired to determine 

 the nitrogen present, on an ash-free filter. 



In either case the precipitate is washed with absolute alcohol two or 

 three times, and is then either dried and weighed, or is transferred 

 along with the filter paper to a Kjeldahl's flask, and the nitrogen 

 determined. 



100 parts creatinin zinc chloride = 6 2 -4 4 parts creatinin. 

 100 parts nitrogen correspond to = 269 parts creatinin. 



The 80 c.c. alcoholic extract employed above correspond to 192 c.c. 

 of the urine employed. By adopting the method described above (i.e. 

 by taking aliquot parts of the various filtrates) the necessity of washing 

 the various precipitates is obviated and much time is thereby saved. 1 



Creatinin also forms a double salt with mercuric chloride having, accord- 

 ing to Johnson, the formula 4(C 4 H r N 3 O . HC1 . HgO)3HgCl 2 . This salt 

 may be obtained by adding to urine a twentieth of its bulk of a cold 

 saturated solution of sodium acetate, and a fourth of its bulk of a cold 

 saturated mercuric chloride solution. An immediate precipitate of in- 

 organic bodies and urates falls down, and if this be filtered off, the filtrate, 

 on standing, gradually develops a further precipitate, said by Jonston 

 and Colls to consist of pure mercuric creatinin. It has been suggested 

 to collect this precipitate, dry and weigh it, and calculate therefrom the 

 amount of creatinin present. On testing the method as above described 

 I found that the results, when compared with those obtained by 

 Salkowski's method, were very unreliable, and that if the first filtrate 

 were warmed as suggested by Halliburton, in order to accelerate the 

 precipitation of the creatinin compound, a black mass of reduced mercuric 

 oxide was almost invariably obtained. Moreover, if the precipitate 

 were collected and dried, and an estimation of its mercury contents 

 made, it was found that the precipitate was far from pure. I found, 

 however, that the impurities were, to a large extent at least, insoluble 

 in 1 per cent. HC1, so that by washing the Johnson's precipitate with 

 acid of this strength, an almost pure solution of the mercury salt in 

 hydrochloric acid was obtained. This solution was then neutralised 

 and its mercury contents estimated by a titrimetric method described 

 by Hannay in Sutton's Volumetric Analysis, p. 241, and from this the 

 Amount of creatinin was calculated. The results thus obtained, when 



1P This method of estimating creatinin although the most accurate is almost 

 prohibitive on account of the expense of absolute alcohol. 



