SECT, vii.] STEAM ENGINES. 219 



places by a top and bottom plate, to which the piston rod is fixed. The segments 

 are pushed away from the centre by steel springs, of the form of the letter V. 



Pistons on Cartwright's plan have not been quite successful in practice, when 

 the cylinders have not been truly bored ; and the causes were pointed out very 

 clearly by Mr. W. Nicholson, soon after the invention was brought before the 

 public. 1 The pieces forming the piston having a determinate curvature, and being 

 too strong to be sensibly flexible, cannot be expected to accommodate themselves to 

 any irregularity in the cylinder in different parts of its length, as is done by the 

 elastic stuffing of hemp ; and there is reason to doubt, in applying them, whether 

 the pressure of the rings or pieces together has not been too powerful for the 

 springs to perform their office when applied in this manner. 



As to the actual difference between the friction of metal, and hemp against 

 metal, when the pistons are equally steam-tight, it is undoubtedly in favour of 

 metallic pistons (art. 463). 



470. BARTON'S PISTON. A piston considered superior to Cartwright's was made 

 by Mr. Barton, Plate vn. Fig. 4. It consists of one thick ring E, of brass or cast 

 iron, made very nearly to fit the cylinder, and then cut into three or more equal 

 segments : the equal triangles remaining are used as wedges to expand the seg- 

 ments of rings into a larger circle. The segments, and small triangles or wedges, 

 are secured between a top and bottom plate, as in the piston last described, with 

 spiral springs to press the triangles outwards from the piston rod, making them act 

 as wedges to press the segments against the inside of the cylinder ; and as these 

 wear by use, the points of the wedges themselves protrude, and, being formed of the 

 same metal, still make part of the piston. A piston of this kind, and a true 

 cylinder, has been known to work for some years without requiring any other 

 attention than keeping it properly greased ; but it is easy to prove that the wedges 

 and segments do not expand equally, hence in this state it was not applicable to 

 high pressures ; besides, the imperfection of Cartwright's piston still remained. It 

 has however been recently much improved by Barton, and therefore I propose to 

 describe it more fully in its improved state. 



The piston is represented by a plan and section, Fig. 4. It is composed of a 

 solid cylindrical cast iron body A, having a conical hole B, to receive the enlarged 

 end of the piston rod, to which it is secured by a cross pin D, passing through 

 both. A space or groove is formed round the body of the piston, to receive four 

 brass, cast iron, or cast steel, hardened and tempered segments marked E, which 

 are spread asunder by four triangular wedges G, of the same metal as the 



1 Philosophical Journal. 



