79. STARCH. Ill 



The process of iDrevious digestion with a more dilute 

 acid separates the starch more completely from the cellu- 

 lose, the former being converted into the soluble modifica- 

 tion, while the latter remains unchanged. {Krocker.) 



h. Wolff's process.— 'D\^Q.^% 2.5 to 4 grms. of the sub- 

 stance with 100 c.c. of water, and 12 to 16 drops of con- 

 centrated acid, 24 hours on the water-bath, then seal the 

 liquid up in glass tubes and heat 12 hours in an oil-bath to 

 120° C, then dilute, decolorize with plumbic acetate, and 

 so on as directed in a, 



3. Dragendorff gives the following method for separat- 

 ing and determining starch and the other matters with 

 which it is usually associated. 



Pulverize about 2.5 grms. of the substance that has been 

 dried at 100°, mix the powder with about 30 grms. of a 

 solution of about 6 parts of j^otassic hydrate in 95 parts of 

 absolute alcohol, and digest the mixture 24 hours at 100° C. 

 in a sealed tube, or in a flask that can be closed air-tight ; 

 filter the contents of the flask, while hot, through a dried 

 and weighed filter, wash the residue thoroughly, first with 

 hot absolute alcohol, then with cold alcohol of ordinary 

 strength, and finally with distilled water, mixed with a 

 little alcohol if gummy substances are present in notable 

 quantity, dry the filter, first at 50°, and then at 100°, and 

 weigh. The difference between this weight and that of 

 the substance taken gives the amount of albuminoids, fat, 

 sugar, and a part of the mineral salts. 



The insoluble residue, with the filter torn in shreds, is 

 heated with water containing 5°|\, of hydrochloric acid 

 until a drop of the liquid gives no blue color with solution 

 of iodine, the mixture is filtered asrain throuo^h a dried 

 and weighed filter, dried at 100°, and v/eighed. The sec- 

 ond loss of Aveight gives the amount of starch ; it was 

 converted into dextrine by the acid and dissolved out ; 

 a very small quantity of mineral matters might j^ass into 

 solution also, and, if great accuracy is required, the 



