66 STUDIES ON APPLES. 



shaken, and allowed to stand over night. The solution was then filtered, the lead 

 nearly all removed by dry sodium sulphate, followed by dry sodium carbonate, the 

 solution again filtered, and the filtrate used in the determination of sugars by polar- 

 imetric ami gravimetric methods. The Clerget method was used for the polari- 

 metric work, calculating by the formula as modified by Tolman, viz: 



141.85 + .062b2- 



Soxhlet's method was used in the determination of reducing sugar as invert before 

 and after inversion. The tables of Meissl and Wein were used, and the cuprous 

 oxid was filtered off on Gooch crucibles and weighed as such, as described by 

 Munson. & 



DETERMINATION OF STARCH. 



Fifty grams of pulp were weighed into a clean cloth bag, the mouth of the bag 

 closed w r ith a rubber band, and the contents squeezed with the hand or lemon 

 squeezer, and then washed with portions of about 25 cc of water until the washings 

 amounted to about 250 cc. The last washings were in all cases neutral to litmus 

 and free from reducing sugar. The starch was settled from the washings by means 

 of a centrifugal machine, repeatedly washed by stirring with fresh portions of water 

 and settling, and transferred to a 300-cc flask, roughly graduated at 200 cc. The 

 marc from the cloth bag, separated as completely as possible by scraping with a 

 spatula, was washed into the flask, and the whole made up to 200 cc. Twenty cubic 

 centimeters of approximately 25 per cent hydrochloric acid (sp. gr. 1.125) was added, 

 and the whole heated for three hours in, not on, the steam bath under an air con- 

 denser. The mixture was then cooled, almost neutralized by sodium hydroxid, 

 cooled again, made up to 300 cc, filtered, and 25 cc portions employed for the 

 determination of dextrose by Allihn's method, c 



DISCUSSION OF METHOD FOR STARCH. 



Determinations by the above method are all 0.5 to 0.8 per cent too 

 high, owing to bodies not starch which remain in the marc and become 

 partially hydro ly zed to reducing bodies by the acid treatment. 



The supernatant liquors from the centrifugal machine always give 

 a test for starch, but the quantity of starch present is very minute. 



Satisfactoiy results could not be obtained by the diastase method,** 

 because* it was impossible to break all the cells. The diastase did not 

 have ready access to the swollen starch grains, and the complete 

 washing out of dextrin was not possible. The direct extraction of the 

 sugars by alcohol was tried, but was found less convenient than the 

 water extraction. 



Other methods for the determination of starch in fruits have been 

 described, among which the methods of Browne and Lindet are given 

 as worthy of special consideration. 



Bui. 73, p. 70, Bureau of Chemistry, U. S. Dept. of Agr. 

 & Ibid, p. 65. 



l. 65, p. 49, Bureau of Chemistry, U. S. Dept. of Agr. 



id, p. 58. 



