16 BUENING QIJALITIES OF TOBACCO. 



residue. A^^iile clilorin is undoubtedly injurious, these experiments 

 all indicated that it requires larger quantities to seriously affect the 

 burning qualities than is commonly supposed. The sulphate, when 

 added in any considerable quantity, invariably injured the burning 

 qualities very markedh^, acting in this respect very much like the 

 chlorid but to a lesser degree. The conclusion reached by Xessler 

 that potassium sulphate is highly beneficial, which was based on ex- 

 periments with filter paper, is thus shown to be erroneous. Potassium 

 nitrate in large quantities causes tobacco to burn explosively and the 

 combustion is incomplete; in smaller quantities it exerts a distinct- 

 ively beneficial action, but not more so than some other potash salts. 

 The quantit}^ of potash combined with nitric acid in tobacco is gen- 

 eralh^ comparatively small, and other forms of potash are more im- 

 portant in promoting the fire-holding capacity. As regards the phos- 

 phates, only the dipotassium salt need be considered, and this appears 

 to be practically neutral in its action, neither promoting nor hinder- 

 ing the fire-holding capacity. Moreover, as compared with the other 

 important acids the quantity of phosphoric acid is nearly always 

 sma^l. The oxalate, citrate, malate, and acetate of potash all showed 

 very beneficial effects in ever}^ case, though much larger quantities 

 were required for some samples of tobacco than for others. Exces- 

 sive amounts of these salts, on the other hand, injured the burning 

 qualities, especially as regards the character of the ash. Also when 

 applied to filter paper in small quantities the latter acquires a good 

 fire-holding capacity, whereas large amounts again destroy this 

 jDroperty. The acetate is considerably^ less efficient in improving the 

 fire-holding capacity than the other three organic salts, probably 

 on account of its greater stability. 



CALCIUM. 



Considerable difficulty is met with in getting accurate tests with 

 the calcium salts because nearly all of them are difficultly soluble, 

 and hence solutions can not be obtained of sufficient strength to give 

 decided results. Only the chlorid, nitrate, and acetate are easily 

 soluble, and of these the nitrate shows an anomalous conduct, and so 

 the results obtained with this salt are not specially significant. In the 

 case of the insoluble salts, emulsions were applied to the tobacco, but 

 of course results obtained in this way are not as reliable as those se- 

 cured by use of solutions. The chlorid of calcium is very injurious 

 to the fire-holding capacity and decidedly more so than the potas- 

 sium salt, so that even small quantities destroy this property. Cal- 

 cium sulphate in moderate quantities injures the burn markedlj^ and 

 to a greater extent than the corresi^onding potassium salt. The effect 

 of the nitrate of calcium on the burn is somewhat surprising. If a 



105 



