IMPROVEMENT OF lU'RNTNG QUALITY. 41 



hiistion under ordinary conditions. Takino; this variability as a 

 I)asis, it has been possiule to produce by careful seed selection strains 

 of tobacco possessing: greatly improved burn without any change in 

 the soil or in the method of handling the soil or the crop. 



In the case of one varietv of Sumatra tobacco to which the jrreat- 

 est objection was its poor burn, strains have l>een produced in the 

 course of these experiments which burn in a perfectly satisfactory 

 way without coaling or ilaking. Even the top leaves in these par- 

 ticular strains have a free, even burn and good capacity for holding 

 fire. 



The production of improved burning strains requires more de- 

 tailed experimental work than the improvement of shape, size, or 

 number of leaves. Xo field character of the plant has been closely 

 enough associated or xjorrelated with the nature of the combustion 

 of the cured leaves to make possil)le the selection of the best burning 

 plants in the field. Consequently they can only be determined by 

 actual burning tests of the tobacco after it has past thru the proc- 

 esses of curing and fermentation. For this reason the leaves of each 

 seed plant must be carefully harvested separately and labeled in a 

 manner to correspond with the label designating the seed saved from 

 the same plant. It is always desirable that each priming of leaves be 

 numbered or marked so that it may be identified after curing 

 and fermentation. This enables the exi)erimenter to make a test of 

 the uniformity of the burn of the top. middle, and bottom leaves of 

 each individual seed plant. There is considerable variability in the 

 degree of uniformity of the burn of leaves borne on diiferent parts 

 of the plant, and therefore it is desirable to secure seed from plants 

 which show a good burn in all the leaves, m order to produce a strain 

 with uniformly good combustion. The leaves of all the seed plants 

 should be cured and fermented under conditions as nearly normal as 

 can be obtained in order to admit of a fair competitive test and to 

 eliminate the possible influence of irregular conditions. 



Where large numbers of samples are to be tested specially con- 

 structed apparatus is necessary to secure accurate results. A simple 

 form of apparatus has been devised by Dr. W. W. Garner, of this 

 office, for making these comparatiA'e tests. It consists of a series of 

 glass tubes so arranged that each tube will smoke a cigar in very 

 much the same manner as it is smoked by an individual, but with 

 more regularity and uniformity. This apparatus is operated by 

 means of an intermittent flow of water which subjects all the cigars 

 to exactly uniform conditions. A carefully adjusted asf)irator draws 

 the proper intermittent current of air thru the cigars, and is so con- 

 nected with the tubes that exactly the same strength of current is 

 drawn thru each cigar. A paper has been prepared bj'^ Doctor Gar- 



96 



