334 EXPERIMENT STATION RECOED. 



determined in centimeters, and with these results as a basis the yield was 

 calculated. 



It was further found that a fair average could be deduced without measuring 

 so many stalks, the difference between the approximate and the true averages 

 of the diameter of the stalks not exceeding i mm. 



In the application of the system the number of cane stalks per acre is esti- 

 mated by counting the stalks in a few of the 24-ft. plant rows, and oy measuring 

 the average height of the stalks at a point where the stalks are erect. In this 

 way it is considered possible to obtain a fairly accurate estimate of the weight 

 of the crop. 



The relation of the composition of the leaf to the burning qualities of 

 tobacco, W. W. Garner (i'. >S'. DciJt. A(/r., Bur. I'Uint Indus. Bui. 105^ pp. 27). — 

 In summarizing the results obtained by different investigators in their studies 

 of the burning qualities of tobacco, it is pointed out that only two facts have 

 not been disputed, the one that ehlorin injures the fire-holding capacity and the 

 other that potash favors this property. These two facts are insufficient to ex- 

 plain the differences in the burning qualities of tobacco, and the author in his 

 experiments worked with a view to determine the effects of sulphates and phos- 

 phates and the relative value of the different salts of potash in promoting the 

 fire-holding capacity. 



The tobacco samples, all of which had been thoroughly fermented, were 

 extracted with various solvents and the effect on the burning qualities noted. 

 In various samples extracted with petroleum ether and with ordinary ethyl or 

 sulphuric ether the burning quality was not appi'eciably affected, and the same 

 result was secured when strong alcohol was used as a solvent, except in one 

 case where the fire-holding capacity was considerabl.v improved by the 

 extraction. 



It is ix)inted out that the active principle or principles imparting fire-holding 

 capacity to the leaf ciin be extracted with \^ter, and that the problem is nar- 

 rowed down to the determination of the composition of the extract and the dis- 

 covery of which of the constituents contribute to the burning qualities. With 

 this object in view, 100 gm. of tobacco having a good burn was extracted with 1 

 liter of distilled water, the extract being poured off and the tobacco again ex- 

 tracted with the same quantity of water for 24 hours longer. The principal 

 constituents of these extracts were the chlorid, sulphate, nitrate, citrate, and 

 malate of potassium, together with ammonium and nicotine salts and small 

 quantities of lime and magnesia. An examination of the ash of the extracted 

 tobacco showed that practically all of the phosphoric acid, about one-half of the 

 magnesia, all of the oxalic acid, and the greater portion of the lime remained 

 in the leaf, while the extract contained nearly all the chlorid, all the potash, 

 and the malic, citric, and nitric acids, and most of the sulphuric acid. About 

 one-half of the total ash was extracted from the leaf by this process, and this 

 seemed to contain all the constituents imparting to the tobacco its fire-holding 

 capacity. 



An extract of tobacco having jioor burning qualities was prepared in the 

 same way. This also showed the iwwer of imparting fire-holding capacity, but 

 as nearly as could be determined this power was ©nly about one-fifth of that 

 of the other extract. The extract from the poorer tobacco contained about 5 

 times as much sulphuric acid, twice as much magnesia, and considerably less 

 nitric acid than the extract from the tobacco with good burning qualities. " The 

 total quantity of iK)tash was about the same in the two extracts, so that the 

 extract from the tobacco with poor burning qualities contained much less potash 

 in combination with the organic acids. The difference in composition of these 

 extracts, then, obtained from tobaccos having good and bad burning qualities. 



