STUDIES ON THE CURING OF LEAF TOBACCO. 9 



The results of the experiment are shown in Table II. While the 

 actual number of leaves in each lot was 58, the results are calculated 

 on a basis of 50 leaves for convenience in comparing with the other 

 experiments. The experiment, of course, does not show the total 

 loss in weight in curing in any case, but other experiments with similar 

 material from the same field indicate that the loss in weight of the 

 picked leaves must have been more than 20 per cent. The type of 

 tobacco and the stage of ripeness account in all cases for the larger 

 losses in weight than were obtained with the cigar types. Since 

 only a single experiment was made and a rather small number of 

 leaves used, the results can be taken as only approximately correct. 

 These results indicate that when the stalk is split in harvesting, the 

 loss in weight in curing is less than when the stalk is not split, but, 

 nevertheless, the loss is much greater than when the leaves are picked 

 from the stalk. 



COMPOSITION OF CIGAR- WRAPPER LEAP BEFORE AND AFTER CURING. 



The material obtained in connection with the data presented in 

 Table I was subjected to analysis to ascertain the nature of the 

 changes in composition which take place in air curing and to which 

 the losses in dry weight are due. Because of the variations in the 

 composition of the leaf at the time of harvesting, it is obviously 

 essential that the material chosen for study shall be obtained under 

 such conditions as will insure original uniformity in the composition 

 of the uncured and cured samples. That this has been very closely 

 attained in the above-mentioned material is shown in the values ob- 

 tained for those constituents which undergo no change in the curing. 



LEAF HARVESTED BY PRIMING. 



The samples 1 used in these experiments cover the curing seasons 

 of 1908 to 1911, inclusive, and include cases in which the curing was 

 not fully completed. Although the midrib is never removed prior to 

 curing, it has little or no value in manufacturing and, moreover, 

 differs decidedly in composition from the leaf web, so that for present 

 purposes it is necessary to consider the two leaf parts separately. 

 The cured leaf always contains more or less water, depending on the 

 type in question and on atmospheric conditions, but to simplify 

 matters all results have been calculated to a water-free basis. 



PREPARATION OF MATERIAL AND METHODS OF ANALYSIS EMPLOYED. 



To pulverize the material for analysis, the leaf web was simply passed 

 through a 60-mesh wire sieve. The stems were prepared by grinding 

 to a fine powder in an iron mortar. The water content of the sample 

 was determined by drying over sulphuric acid, as recommended by 



1 For details as to the methods followed in selecting and gathering the samples, the type of leaf used, etc., 

 seep. 3. 



29731°— 14 2 



