Nutrition Investigations. 311 



This was easily filtered off through gauze, as much liquid as possible 

 removed by pressure, and the precipitate washed quickly with a 

 little water to remove the excess of alkali. The carbohydrate was 

 freed from its copper compound just as described by Salkowski (140). 

 The precipitate was placed in a mortar and rubbed to a cream with 

 diluted hydrochloric acid (1 volume of water to 1 volume of concen- 

 trated acid) the acid being added until all blue particles had disap- 

 peared. It was then poured into 90 per cent alcohol, the precipitate 

 filtered off upon plaited paper and washed with 50 per cent alcohol, 

 replaced in 90 per cent alcohol acidified with hydrochloric acid, and 

 allowed to stand several hours to dissolve out the copper. It was then 

 filtered, dissolved in dilute potassium hydroxide, and the dark brown, 

 muddy solution filtered through paper on a hot funnel, the carbohy- 

 drate reprecipitated with acid alcohol, and redissolved and reprecipi- 

 tated until free from copper. When it no longer came down readily 

 in alcohol, acetone was substituted, in which it formed white fibrous 

 masses resembling paper pulp. Washed with absolute alcohol and 

 ether, and dried in vacuo over sulphuric acid, it became a cream-white 

 powder. Both of these methods yielded a product readily soluble in 

 cold water, forming a clear, limpid, amber-colored solution. It gave 

 no color reaction with iodine, and contained no reducing substance. 

 In Fehling's solution it formed a very flocculent white precipitate, was 

 not precipitable by lead acetate, neutral or basic, in neutral solution, 

 but formed a precipitate in alkaline solutions. A test for mucic acid 

 gave negative results, but a strong furfurol reaction was obtained on 

 boiling with hydrochloric acid, indicating the presence of pentosans. 

 A 1-gram sample of material, prepared by the method first described, 

 was tested quantitatively for pentosans. It contained 26.8 per cent 

 moisture, and 2.48 per cent ash, and yielded 0.076 grams of phloro- 

 glucid, from which the yield of pentosans, according to Krober's 

 tables,! is calculated as 72 per cent. The phoroglucid precipitates were 

 afterwards extracted with 95 per cent alcohol, according to Ellett 

 and Tollen's^ method for quantitative determination of methyl-fur- 

 furol. The Gooch crucibles containing the precipitates were warmed 

 10 minutes to 60° C. with 15-20 cc. of alcohol, the extract filtered off, 

 and the extraction repeated till the alcohol was colorless. The pre- 

 cipitates were then dried at 100° C. and weighed. The loss of weight 

 was 0.0047 grams or 6 per cent of the original precipitate. The dulse 

 preparation therefore contained a small amount of methyl-pentosan 



iZeitschrift fiir physiologische Chemie, XXXVI, appendix. 



^Berichte der deutschen chemischen Gesellschaft, Vol. 38, p. 492 (1905). 



