480 Journal of Agricultural Research voi. xix, no. 10 



The whole sample was then run through a sampling press, constructed 

 on the principle described by Clark (16). In this press the tissue was forced 

 through fine perforations which left it in a finely divided state. It was 

 then very thoroughly mixed by stirring and carefully weighed out in 

 six 50-gm. portions. By this method there is a very slight loss in moisture 

 due to evaporation; but by working as rapidly as possible after the 

 sample is run through the press this loss is reduced to a minimum and is 

 closely comparable in all cases. The method of sampling employed was 

 very satisfactory from the standpoint of giving close checks on such 

 duplications as were run. 



Sugars and acid-hydrolyzable reducing material. — Duplicate 

 samples were prepared for these determinations, but only one was 

 run through. The 50-gm. samples were covered with about 100 cc. 

 of 95 per cent alcohol, 3 to 5 drops of ammonia were added to neu- 

 tralize the acid, then the samples were immediately put on the steam 

 bath and boiled vigorously for a few minutes. Most of the excess am- 

 monia was driven off by the boiling. Preliminary tests of glucose solu- 

 tion treated in this manner \vith dilute ammonia showed no measurable 

 breaking down of the sugar. The alcohol was then filtered off through 

 an alundum thimble, and the sample was extracted with alcohol in a 

 Soxhlet extractor for about 12 hours. The extract was then added to 

 the original alcohol filtrate, and the sample was made up to 500 cc. with 

 distilled water. Fifty cc. of this were cleared with neutral lead acetate 

 and were made up to 250 cc. The excess lead was removed, and the 

 sugars were determined in 20 cc. of the cleared solution, according to 

 Mathews's^ modification of Munsen and Walker's and Bertrand's methods. 

 Duplicate titrations with permanganate solution were checked to within 

 0.2 cc. 



Fifty cc. of the cleared solution were inverted in 3.5 per cent hydro- 

 chloric-acid solution, standing about 20 hours at room temperature. 

 The total reducing substances were then determined in this solution. 



The residue in the thimble from the Soxhlet extraction was re- 

 moved quantitatively, was dried, and 125 cc. of 2.5 per cent 

 hydrochloric acid were added for hydrolysis by the modified Sachsse 

 method.^ Reducing substances were determined as outlined above 

 and were figured as dextrose. This alcohol-insoluble acid-hydrolizable 

 material probably consists mainly of starch, galactans, hemi-celluloses, 

 and pectin in the green fruit fresh from the tree. In the ripe fruit 

 there is no starch present, as shown by the iodin test, and the reduc- 

 ing material must come almost entirely from such sources as the 

 galactans, hemicellulosses, and pectins. Much of the reducing material 



' Mathews, A. P., physiological chemistry. . -ed. a, p. 994. New York, 1916. 



2 Wiley, H. W., ed. offioal and provisional methods of analysis, association of offioal agri- 

 cultural CHEMISTS. As compiled by the committee on revision of methods. U. S. Dept. Agr. Bur. 

 Chem. Bui. 107 (rev.), p. 53. 1908. 



