24-8 On an Improvement in the Apparatus for procuringPotassitmu 



operating which I pursued might not be wholly unworthy the 

 notice of this Society. 



It is well known that the grand difficulty in successfully per- 

 forming the experiment in question, consists in protecting the 

 gun-barrel from the effects of that extreme and long-continued 

 heat, which is necessary to decompound the alkali, and to vo- 

 latilize its base. The usual practice hitherto has been to sur- 

 round with a lute *, that portion of the gun-barrel which is 

 introduced into the fire. This operation, however, is always 

 tedious ; and although it be conducted even with the greatest 

 care, it is found extremely difficult to prevent fissures in the 

 coating, particularly when the heat is much increased in the 

 course of the experiment. Hence, if eventually the fire have 

 direct access to the barrel, through any crevice which may be 

 fonned, the fiision of the denuded part is generally the con- 

 sequence, and the whole labour of the experiment is lost. 



This, then, being the common cause of failure, it occurred 

 to me that, if there were substituted for the luting, a thin but 

 sound and well-burnt tube or hollow a/lmder of Stourbridge 

 clay, of such dimensions as just to cover that portion of the 

 barrel which is subjected to the fire, the unfortunate result, 

 which I have alluded to, might possibly be avoided. 



A tube of this description was accordingly procured ; and, 

 in order to guard against the hazard of its cracking, by reason 

 of exposure to a sudden increase of temperature, it was, in the 

 first place, gradually and with caution, heated to redness. 



The remaining part of the experiment was then performed 

 with entire success; and a veiy considerable quantity of potas- 

 sium obtained. 



It may be proper to remark that the hollow cylinder, and 

 tliat portion of the gun-barrel which it incloses, should be of 

 such relative diameters that, when cool, their corresponding 

 surfaces are not quite in close contact ; otherwise, the cylinder 

 will be in danger of bursting, not only on account of its own con- 

 traction, but also on account of the simultaneous expansion of 

 the gun-barrel, fi-om the effects of that very high temperature, 

 to which, in this state of combination, they are submitted. 



Moreover, the whole apparatus should be supported accu- 

 rately in the same position, thi'oughout the experiment, (by 

 means of rests made of Stourbridge clay,) and should be so 



* " On couvre cette partie raoyenne d'un lut infusible." — Gay-Lussac et 

 Thenard, Ann, de C/iim. torn. Ixvi. p. 20/. 



" The Ittfc which I have found most effectual was composed 



of Stourbridge clay &c."— Tennant, I'hil. Trans, for 1814, p. .582. 



" It (the gun-barrel) is then covered with an uifusible /«/<■."— Brande'g 

 Manual of Chemistry, p. 184. Ed. 1819. 



situate 



i 



