Dcscr'iplion of a Safety Falue. 255 



fixed in the boiler cd, having four openings, ii, which are 

 represented in a plan view in fig. 2. ef is the metaliie rod, 

 bearing the weiizht KK, with which the Safety valve is 

 loaded, and extending itself under that valve to /': g/i is 

 the vacuum valve, consisting in a plane circular plate, with 

 a brass tube sliding along the rod, and pressed by a spiral 

 spring to the safetv valve ah (against which it has been 

 well ground in making it), closuig in that situation the 

 openings ii. 



Such being the construction of the whole, it is evident, 

 that when the elasticity of the steam increases, the two 

 valves, joined together, with the holes ii shut, make but 

 one, opposing to the elasticity of the steam an united re- 

 sistance, which is regulated by the weight kk, in the com- 

 mon way ; but, on the contrary, \^ hen by condensation of 

 the vapours a vacuum is produced, the external air in press- 

 ing through ii, upon the vacuum valve g/i, forces it down, 

 and opens to itself a passatje into the boiler. 



The valve gh mav easilv be made conical, like the other, 

 if that form should be preferred ; but in different trials I 

 liave found plants, if well turned and ground together, join 

 as perfectly as can be desired, being pressed by the united 

 elasticity of the spring and the steam. 



Fig. 3 is the same contrivance adapted to a new kind of 

 safety valve or piston, which, though I ori<iinally intended 

 it for the use of Papin's digesters cA a new construction *, 

 has been, in a larger size, applied b)' me even to steam en^ 

 gincs, and is described in the Pliilosophical Magazine of 

 December i &03 f. 



I have lately begun, and shall pursue, a set of experi- 

 ments, with the intention of regulating by this safety piston 

 the quantity of admitted air to lire-grates, and lo effect by 

 that means a new mode of regulating the fire, and the elas- 



* Nicholson's Journal, March 1804. 



t The description of this contrivance being already published, it would 

 be superfluous to repeat it. 1 (jn!y beg leave to add the following prac- 

 tical remark : — .A metallic piston, if well turned and fitted into a cy- 

 linder of exactly the same kind of metal, will probably have the same 

 degree of expansion, tsptcirilly if hollow, and consccjucntly will not in- 

 crease its friction in any increased degree 01 temperature. But as m 

 practice the cylinder is commonly exposed to a lower temperature than 

 the piston, heated t>y the stearu, a little increase of friction will take ph.tce 

 by an increase of heat. 1"o prevent the effect of this, I have found it 

 Useful to employ for the piston a metal of somewhat less expansive poweis 

 than the cylinder; and the e.xp insion of red copper being to that of bras'. 

 nearly as 10 to ii, J prefer rnakiiijj ih.c pi>ton o; the turmcx metal when 

 the cylinder is niuJe of bra^s. 



ticity 



