Of the DEVON IRON WORKS. 41 



grefs in this matter, by laying me under the neceflity, two years 

 ago, of withdrawing myfelf entirely from the concern. 



I have the honour to be, refpedfully, 

 SIR, 



Your mofl obedient fervant, 

 EDi^fBURGH, -1 ^oj^^ Roebuck. 



June 30. 1798. J 



T'o Sir 'James Hall, Bart. 



In order to illuftrate what is faid above, a ground plan of the 

 air vault and furnaces of the Devon Iron Works is given in 

 Tlate I. ; of which the explanation follows : 



Explanation of Fig. 2. Pl. I. 



A The air vault, formed by a mine drove in the folid 



rock of coarfe grained freellone. 



B The blowing cyhnder. 



C The pipe that conveys the air from the blowing cy- 



linder to the air vault. 



D The edudion pipe that carries the air from the air 



vault to the iron wind chefl. 

 Part. I. F E The 



AGAiJf, this quantity multiplied into its velocity, will be as the momentum of 

 the fluid expelled, or as the power by which it is expelled, that is, V^D^ = P, or 

 VD = /P. 



Here, therefore, if D is given, V is as \/P, as Mr Roebuck affirms. Alfo, 



O /P 



becaufe V = ^, and alfo V = ^-jr, (^r= Dv/P» fo that, while P remains the 



fame, Q_will increafe as D increafes, and V will diminifli in the fame ratio. 



The problem, therefore, of throwing the greatefl quantity of air into the furnace, 

 with a given power, ftriftly fpeaking, has no maximum, but the largeft aperture of 

 which the engine can admit mufl; be the beft. It is probable, however, that there 

 is a certain velocity with which the air ought to enter into the furnace ; this will 

 produce a limitation of the problem, which, as Mr Roebuck fuggefls, is not likely 

 to be difcovered but by experiment, J. P. 



