16 



from 4 to 5 per cent. According to Jenkins/ the losses may be even 

 larger. He reports that "the upper leaves, short seconds, and first 

 wrappers lost, respectively, by fermentation 9.7, 12.3, and 9.1 per cent of 

 their total weight. But while three-fourths of the loss in the case of 

 the short seconds consisted of water, in the case of the upper leaves 

 almost three-fourths of the loss was dry matter. The first wrappers 

 lost a little less dry matter than water." 



Development of gases. — The formation of ammonia can be noticed by 

 the characteristic odor in the fermenting rooms, but the amount is not 

 so high as one might naturally be led to suppose from the intensity of 

 the smell. About 3 liters of air from the interior of a fermenting pile 

 when drawn through 25 cc. of Nessler's reagent, produced a light yellow 

 color, indicating about 0.05 milligram of ammonia. No trace of hydro- 

 gensulphide is given oft'. Test tubes containing filter jiaper moistened 

 with basic lead acetate remained perfectly colorless for twenty-four 

 hours in the fermenting heaps, hence it may be safely concluded that 

 no protein decomposition resembling putrefaction takes place.^ The 

 amount of carbonic acid given off was also much smaller than would 

 naturally be exx^ected from the apparent energy of the action. 



Starch. — Small quantities of starch are sometimes found in fer- 

 mented tobacco when the curing process has not been carried out proiJ- 

 erly in all parts of the leaf or in x^arts of leaves broken, or injured by 

 fungi, as observed by Miiller-Thurgau, but this occurrence of a small 

 percentage of starch interferes with the flavor just as little as does the 

 closely related cellulose. The Avell-prepared tobacco wrappers from 

 Florida examined by the writer did not show a trace of starch. The fact, 

 however, that in the curing process the solution of the starch is going 

 on with great energy forms a contrast to the observation that in cer- 

 tain cases remnants of starch remain unattacked during the fermenta- 

 tion process. This admits of hardly any doubt that the diastase^ is 

 gradual!}^ destroyed, perhaps by the proteolytic enzym. 



Sugar. — As to the disappearance of the last remnant of sugar during 

 the sweating, amounting, according to Miiller-Thurgau, to from 1 to 3.3 

 l)er cent, some authors assume oxidation to carbon dioxide and water, 

 and others assume a i^artial transformation to acetic acid. When it is 

 taken into consideration that an alkaline medium can soon change 

 glucose into organic acids (gluconic, saccbarinic, etc.), especially in 

 the presence of air, a more simple explanation would be at hand than 



1 Conn. AgT. Expt. Sta., Ann. Rept. 1892, p. 28. The leaves used for comparison were 

 most carefully selected and were as nearly alike in color, size, and texture as pos- 

 sible. 



^Nessler's c^-nparison of the sweating process to putrefaction is certainly not 

 admissible • nei >her is his declaration that the formation of ammonia is not normal, 

 but simply a sign of true putrefaction. 



^Diastase is absolutely necessary to dissolve and saccharify the starch. The 

 dextrin and maltose thus formed may afterwards be transformed into glucose by the 

 living protoplasm itself, wherever this latter comes under consideration. 



