where the percentage of carbon is so low. Under ordinary conditions it 

 is impossible to obtain a complete combust'on — the ash is usually more 

 or less dark colored. Some authors recommend mixing the substance 

 intimately with fine copper oxide. This procedure is very serviceable 

 when burning salts of these organic phosphoric acids with organic bases, 

 like the strychnine salts which we have previously reported, but with 

 barium salts we have not found copper oxide to be of much use. In the 

 analyses reported above we have used the following method, for the 

 suggestion of which we are indebted to Prof. E. Fischer of Berlin. 



The substance is first burned in the usual way in a current of 

 oxygen, the combustion lasting about an hour. The calcium chloride tube 

 and the potash bulb are then weighed. The increase in weight of the 

 calcium chloride tube is taken as the correct weight of the water. The 

 residue in the boat, which is dark colored from particles of unburned 

 carbon, is powdered in an agate mortar with some recently fused potas- 

 sium bichromate and again placed in the boat, the mortar being rinsed 

 out with some more powdered bichromate. The whole is again burned 

 in the usual way. The potassium bichromate fuses and oxidizes all the 

 carbon in the residue. The increase in weight in the potash bulb is 

 added to the first, giving the total carbon dioxide. 



The analytical data of these various crystalline barium salts, as well 

 as of the free acids from them, is reproduced below in tabular form. 

 These compounds represent the purest preparations that we have been 

 able to produce. 



The relation of carbon to phosphorus in all of these salts is very 

 closely that of 1 : 1 or 6 : 6, and the composition agrees fairly closely with 

 calculated percentages for salts of inosite hexaphosphoric acid. We 

 were forced to the conclusion, therefore, that phytin or inosite phos- 

 phoric acid existing in the natural plant material which we have examined 

 is represented by the formula C„H 18 24 P 6 or inosite hexaphosphoric acid. 

 In the beginning of our study we assumed that the old formula of Post- 

 ernak, C,H 8 O n P,, or as formulated by Neuberg, C(.H., 4 0., 7 P G , was correct. 

 The analysis of amorphous and less carefully purified material gives 

 results in close agreement with this formula. In the carefully purified 

 re-crystallized barium salts, which we have since prepared and which 

 have been described above, the composition is slightly different and the 

 results agree more closely with the phytic acid formula of Neuberg 

 minus 3 molecules of water; viz., C 6 H., 4 O 27 P — 3H 2 = C H 18 O 24 P c . In 

 the analysis of the free acid which we have made we have always found 

 that the hydrogen is too high and the phosphorus too low, in some cases 

 one per cent, too low. The free acid is not suitable for analytical de- 

 terminations for, as we have shown, it can not be dried for analysis 

 without undergoing considerable hydrolysis. With the hydrolysis is 

 associated the absorption of water and production of free phosphoric 



30 



