[ 109 ] 



iron cylinders be objeded to as dangerous, they may be fhod 

 with brafs, which will be found fufliciently hard for this purpofe. 

 I mnft, however, obferve, that the former are already ufed in 

 fevcral mills ; and intelligent powder-makers allow, that acci- 

 dental explofions are moft frequently produced by the collifion 

 of chips which break from the edges of ftone rollers. I am 

 aware of one plaufiblc objedion to fluted cylinders ; the pafte, 

 if very moift, may adhere to the grooves ; but this I think will 

 be prevented by the application of oil to the fluted furfaces in 

 fuch fmall quantities as fhall not injure the compofition. 



Before I difmifs this part of my fubjed I will hazard pro- 

 pofing another alteration in the conflrudion of powder-mills j it 

 is fimply working four rollers in the fame trough inftead of two. 

 Where water is the moving power, the cofi of additional mecha- 

 nifm will appear trifling, if oppofed to the time and labour 

 obvioufly faved by the adoption of this idea, 



I WILL not prolong a tedious, and I fear a tirefome difler- 

 tation, by introducing a minute detail of the procefl^es of granu- 

 lating and drying powder, but content myfelf with obferving 

 that the firft is performed by a horizontal wheel, on which are 

 fixed circular fieves, with parchment bottoms, perforated to the 

 largeft intended fize of the grain ; in thofe fieves the pafle is 

 dcpofited, and with it (in each of them) a fmall oblate fpherical 

 piece of lignum-vitas, which being moved about the fieve by the 

 adion of the wheel, breaks the compofition, and forces it 

 through the parchment bottom into veflTels placed for its recep- 

 tion ; 



