50 The Philippine Journal of Science lais 



suggested by Jacobs.^ This, however, afforded no particular 

 advantage, as the vitamine apparently distributed itself more 

 or less equally between the amyl alcohol and water layers. It 

 was also discovered that the phosphotungstate of the vitamine 

 dissolved fairly readily in an excess of phosphotungstic acid, 

 thus enabling it to be separated from the bulk of the precipi- 

 tated phosphotungstates. But since barium hydroxide was the 

 only reagent found which could be used effectively for decom- 

 posing the separated phosphotungstates, this method was also 

 abandoned. 



When it was desired to isolate considerable quantities of 

 vitamine in a fairly pure or at least highly concentrated con- 

 dition, it was necessary to fall back on the method of Funk in 

 practically its original form as nearly as I now understand it. 

 This was the method used for the treatment of the human cases 

 recently reported by Doctor Saleeby and me.^ 



Every resource has been exhausted to avoid mixed deposits 

 from the final liquor and to increase the yield of needles melt- 

 ing at 233°, but without avail. Every step in the process has 

 been modified in turn, but without materially improving the 

 final result. Among such modifications may be mentioned the 

 following : 



1. Varying degrees of expression of the original polishings. 



2. The use of neutral alcohol for extraction and subsequent hydro- 



lysis of the aqueous extract with dilute sulphuric acid. 



3. The use of a glass still and high vacuum for concentrating the 



extract. 



4. More or less prolonged maceration of the phosphotungstates with 



barium hydroxide. 



5. Removal of ammonia from the resulting solution with a stream 



of air. 



6. Varying the strength of reagents and solvents at many points in 



the process. 



I was, therefore, at a loss for an adequate explanation of 

 the discrepancies between my results and those of Funk. Recent 

 work^ now shows that Funk was in error in reporting the 

 isolation of the curative substance as a pure chemical individual 

 and that the substance melting at 233° consisted almost entirely 

 of nicotinic acid. A number of other differences exist between 

 Funk's results and my own, which are not as yet wholly ac- 

 counted for. I have noted that the quality of the rice 



Woum. Biol. Chem. (1912), 12, 429. 



*This Journal, Sec. B (1915), 10, 99. 



*Drummond and Funk, Biochem. Joum. (1914), 8, 598. 



