290 Mr. Gr. S. Johnson. 



failed to obtain this salt after the use of baryta-water, as directed by 

 Liebig. 



But the action of baryta- water on the watery solution of KH 2 P0 4 

 would be represented thus 



2KH 2 P0 4 + 3Ba(HO) 2 = Ba 3 (P0 4 ) 2 + 2KHO+4H 2 0. 



And the potassic hydrate thus produced would of course decompose 

 the hydrochlorides of any organic bases present in the solution, 

 forming potassium chloride and liberating the bases. The KC1, 

 therefore, would be a product, not an educt, from the flesh. 



Experiment II. This was a preliminary experiment made with 

 50 Ib. of butcher's beef, in order to ascertain the action of 'mercuric 

 chloride in aqueous solution. Briefly summed up, the results were as 

 follows : ^ 



The addition of a sufficient quantity of mercuric chloride solution 

 to the watery extract of fresh beef causes complete and instantane- 

 ous precipitation of the albuminoid constituents and of the whole of 

 the colouring matter. 



The filtrate from the coloured albuminous precipitate deposits, on 

 standing, a spherical mercury salt, isomorphous with the spherical 

 mercury salt of urinary kreatinin described by me in the 'Roy. Soc. 

 Proc.,' vol. 43, pp. 493534. 



This spherical mercury salt yields a kreatinin isomorphous with 

 the tabular kreatinin of urine, when subjected to similar treatment to 

 that described in my paper quoted above. 



Encouraged by these results, I determined to apply the mercuric 

 chloride method, which I had formerly made use of in examining the 

 kreatinin of urine, to the examination of the bases in the juice of 

 flesh. 



2nd. Experiments Designed to Exclude Errors due to Bacterial Action, 

 as well as those due to Chemical and Physical Action. 



Before describing these experiments, I will briefly indicate some of 

 the special advantages possessed by what I have called the " mer- 

 curic chloride method " for the examination of organic bases. 



The method itself, as applied to the examination of urine, is fully 

 described in my paper quoted above. It enabled me to isolate from 

 human urine what I believe to be the excrementitious kreatinin 

 which has always been secreted by the human kidney, but whose 

 properties I was the first to describe. 



Perhaps the claims of my method to excellence in preparing an 

 educt will be best appreciated by comparison with other methods, as 

 in the following table, 



