116 Mr. Nicholson on the Composition of Caff em, 



and also to an analysis which M. Martins * made of guaranin, 

 a substance, the identity of which with cafFein and thein had 

 previously been pointed out by Berthemot and Dechastelusf. 

 Lately Dr. Stenhouse :j:, when examining Paraguay tea, has 

 also made some analyses of thein. 



The following table, in which I have recalculated these ana- 

 lyses according to the atomic weights, carbon 6 and hy- 

 drogen 1, allows a comparison to be made of the results ob- 

 tained by these chemists. 



Mean of the Analyses. 

 CafFein. Thein. Guaranin. 



-> 



Liebig & Pfaff. Wohler. Mulder. Jobst. Stenhouse. Martius, 

 Carbon » 49-30 49-25 49-18 49-47 48-95 49-23 

 Hydrogen 5*22 5-43 5-49 5-20 5*15 5-08 

 Nitrogen. 28-86 28-53 28-90 28-83 



The most simple expression which can be deduced from 

 these numbers is 



Cs Hg Ns O^. 

 Stenhouse's analysis however of the platinum compound 

 proves that this formula must be doubled, and that the atom 

 of cafFein or thein is 



CjsH.oN^O^. 

 The theoretical numbers of this formula are the following: — 

 16 eqs. Carbon ... 96 49-48 



10 ... Hydrogen . . 10 5-15 



4 ... Nitrogen . . 56 28-86 



4 ... Oxygen . . . _32 16-51 



] 94 100-00 



From these numerous experiments the composition of caf- 

 fein might have been considered as perfectly established. In 

 a recent investigation of coffee, however, M. Payen§ states 

 that he has obtained results which differ very sensibly from 

 those obtained by his predecessors, and which he has trans- 

 lated into the formula 



which contains 1 equiv. of oxygen less than the formula up 

 to the present time admitted. 



The theoretical numbers of Payen's formula are — 

 16 eqs. Carbon . . . 96 51*43 



10 ... Hydrogen . . 10 5*35 



4 ... Nitrogen . . 56 30-34 



3 ... Oxygen . . . ^ 12-88 



186 100-00 



* Liebig's Annalen, xxxvi. 93. t Ibid, xxxvi. 90. 



% Mem. Chem. Soc, vol. i. pp. 215, 237. [Phil. Mag., xxiii. p. 426.] 

 § Compies Rendus de I'Academie, tome xxiii. 8. 



