308 Mary Davies Swartz, 



over steam radiators. The Irish moss, being comparatively free from 

 salt, etc., and largely soluble in pure water, was quickly washed once, 

 and extracted immediately. 



All hydrolyses of carliohydiatts were made with 2 per cent hydro- 

 chloric acid, by boiling with a reflux condenser over a free flame. 

 After cooling, the acid was neutral' z d with potassium hydroxide, 

 using phenolphthalein as an indicator, when the solutions were suf- 

 ficiently light in color; in other cases, litmus paper was employed. 

 When the products of hydrolysis served to determine the nature of 

 the carbohydrates, they were evaporated on a water bath nearly to 

 dryness, the residues extracted with hot 95 per cent alcohol the alcohol 

 removed from the filtered solution by evaporation, the residues fre- 

 quently taken up in a little water and decolorized with charcoal, con- 

 centrated, and again extracted with absolute alcohol. 



All qualitative tests for reducing sugar were made with Fehling's 

 solution; all quantitative tests by AUihn's gravimetric method for 

 dextrose, the results being calculated as dextrose in view of the com- 

 plex nature of most of the products, and the advantage of uniformity. 

 On all preparations used for feeding experiments, the length of time 

 in which the maximum yield of sugar could be obtained has been de- 

 termined, as a criterion in analyses of faeces. Five grams of dry air 

 material were hydrolyzed in 500 cc. of 2 per cent hydrochloric acid, 

 50 cc. being removed at intervals of one or more hours, cooled, neu- 

 tralized, made up to 100 cc. and reducing power determined as dex- 

 trose by Allihn's gravimetric method. 



Tests for the presence of fermenting sugars have been made in 

 fermentation tubes with fresh compressed yeast, using as controls 

 solutions of the substance to be tested, without yeast, and dextrose 

 solutions with yeast. 



All carbohydrate solutions for polariscopic examination have been 

 clarified by addition of an equal volume of alumina cream. 



Qualitative tests for pentosans have been made by boiling the sub- 

 stance to be tested in a small Erlenmyer flask with 12 per cent hydro- 

 chloric acid and testing for furfurol with anilin-acetate paper. 



Quantitative tests for pentosans have been made by the furfurol- 

 phloroglucin method. ^ 



Tests for galactans or galactose have been made by oxidation with 

 nitric acid to mucic acid, and the mucic acid identified by !ts melting 

 point (212° C.-215° C). 



'Described in "Official and Provisional Methods of Analysis," United States 

 Department of Agriculture, Bureau of Chemistry, Bull. No. 107, 1907. 



