15 



fermenting piles and that undue pressure must be avoided in order not 

 to diminish this access of air more than is necessary to insure an accu- 

 mulation of heat. Not only are numerous little channels left naturally 

 in the piles, but diffusion also will set in as soon as the air in the piles 

 becomes warmer than the surrounding atmosphere. 



Eepeated efforts have been made to replace the sweating or fermen- 

 tation i^rocess by a direct oxidation. Dr. Mew, of the Army Medical 

 IVIuseum of this city, assures the writer that some experiments made by 

 him about twenty years ago to improve the cured tobacco leaf by direct 

 application of a dilute solution of permanganate of i^otassium resulted 

 in an essential improvement, the i)roduct being milder. Similar results 

 have been recently mentioned by Kiessling.^ In Germany a patent 

 has been granted to the firm of Siemens & Halske for treating tobacco 

 with ozone. However, oxidation often takes quite an undesirable turn, 

 and the danger of destroying the aroma is quite as great, if not greater, 

 than the likelihood of developing it by artificial means. 



Losses. — Jenkins has shown ^ that the losses in fermentation are 

 apparent in the nicotine, protein, amido compounds, nitrogen-free 

 extract, and also, to a much less extent, in the ether extract. The 

 loss of nicotine varies considerably in different sami)les and was 

 found by Jenkins to range from one-sixth to one-half in three samples 

 analyzed. Behrens observed in one sample a decrease of nicotine 

 from 1.46 per cent in the cured leaf to 1.07 per cent in the fermented 

 leaf. Dambergis found in air-dry Greek tobaccos, having from 7 to 14 

 per cent of water, from 2.8 to 0.7 per cent of nicotine.^ 



The question as to how much the loss of organic matter amounts to 

 during the sweating i^rocess can be answered only approximately and 

 by comparing parts of one and the same leaf, but a constant result 

 will never be reached, as the nature of the proceeding in fermentation 

 brings on differences in temperature, water content, and access of 

 oxygen, and thus leads to variations. In the fermenting heap thick 

 and thin leaves occur, often varying more than 20 per cent in weight 

 for an equal surface area. Leaves grown in the shade are thinner than 

 those exx)osed to direct sunlight, and in hot, dry summers the leaves are 

 thicker and coarser than in moist, rainy seasons.^ These conditions of 

 course naturally influence the result. 



Some tobacco manufacturers estimate the average loss during the 

 fermentation process to be 15 per cent (organic matter and water 

 together), while others estimate the loss of solid matter alone to be 



1 Der Tabak, Berlin, 1893. 



^ Conn. Agr. Expt. Sta. Ann, Rept., 1891. 



3 Oesterreich. Cliem. Zeitg., No. 16, 1898. 



^ In the rainy season of 1891 Sumatra tobacco leaves weiglied 52 grams per square 

 meter, while in the dry season of 1892 the leaves grown on the same spot weighed 

 80 to 90 grams per square meter. Behrens explains this difference by the larger 

 intramolecular sj^aces produced by excess of moisture (Landw. Vers. Stat., 1894, 

 Baud 43, p. 272). 



