Proceedings of the Polytechnic Association. 837 



remainder represents the force available to produce motion. From 

 this, however, is deducted the friction of the load, usually called seven 

 and a half per cent; and tlienet power is calculated from the second 

 remainder. For instance, if the mean working pressure is forty-two 

 pounds, and the friction pressure two pounds, forty pounds is avail- 

 able to produce motion without a load ; and seven and one half per 

 cent of this, or three pounds, represents the friction of the load ; so 

 that five pounds pressure is lost in friction, or about twelve per 

 cent of the whole. This mode of calculation cannot always be 

 depended upon. We have known a case where the mean indi- 

 cated working pressure in the cylinder was only eight pounds, 

 and the friction pressure two pounds. Consequently, by the above 

 method, about thirty per cent of the power was absorbed by 

 friction ; but the dynamometer showed that less than ten per cent 

 was lost in that way. Similar cases, differing only in extent, Avill be 

 found quite frequent. The reason is, that engines are packed for the 

 working, and not for the friction pressure. If the steam pressure be 

 100 pounds, the packing must embrace the piston and valve rods with 

 sufficient force to prevent leakage, or say 105 pounds for every square 

 inch of surface packed ; and nearly tlie whole of this will produce 

 friction, when a low pressure is used, but the full pressure will work 

 in between the surfaces, and force back the packing during the steam 

 stroke, so that the friction from that source will be least when the 

 engine is doing its regular duty. Spring packed pistons modify the 

 friction in the same way. In very large engines the state of the pack- 

 ing would have little iniiuence on the friction, though it certainly 

 w^ould seem proper to loosen the stuffing boxes before taking friction 

 diagrams. In some cases, engines are so weakly constructed, that, 

 though the indicator ma}^ shew little friction, without a load, there 

 will really be a great loss Avhen the work is being done, due to parts 

 springing out of line, etc. The dynamometer furnishes, therefore, 

 the only true means of obtaining the net power. In well constructed 

 engines we should be able to calculate the friction by regarding the 

 weight of the moving parts as part of the load, which is moving with 

 a certain velocity in bearings of a given material, and having 

 therefore a co-efficient of friction varying between five and eight 

 per cent. For ordinary purposes, when trial is not convenient, 

 we may assume the friction of small engines, of bad design, or of 

 any engine with weak framing, as being from twenty to twenty- 

 five per cent of tlie indicated power ; wdiile in good engines, of ordi- 



