STATIONARY ENGINES.

215

simple. It was substituted for the double-beat valve in use, andimmediately upon starting it lightened the bürden of the enginesabout 168 lbs., and has since given great satisfaction. This valvepossesses other advantages. In the first place it is much safer thanany other form of valve, as will be easily seen. Supposing a pieceof wood or other material should pass through the pump, as isfrequently the case: if the wood should be caught on the beat ofthe ordinary valve, it would hold the whole valve open and let theengine “come out” with a run, possibly causing considerable damage;but with the small balls of this valve a piece of wood so caughtcould only affect one out of fifty-six balls—so small a percentageof the whole opening that it would merely enable the man in chargeto perceive that there was some trifle amiss by an increase of leak-age. In the second place, the balls, being nearly of the samespecific gravity as the water, are floated open the moment the cur-rent turns in their favour; whereas in all other valves, in additionto the column of water to be lifted, there is also the weight of theheavy metal valve to be opened and held suspended during eachstroke. With large valves this point becomes one of great import-ance, as they often weigh 5 to 6 cwts. each, Again, in this valve,whilst the area of discharge may be made fully equal to that of theplunger, the area exposed to concussive action in closing is reducedto the smallest possible limits, being practically the impinging forceupon one ball, the last one that shuts, or 5-Vth part of the total areaof beating surface; this is owing to the fact that the balls do notall rise to the same height above their seats, and consequently, asthe force of the current acts upon each ball separately, on thecessation of motion each shuts in accordance with the height it hasto fall, and a communication exists between the water on the upperand under side of the valve until the absolute closing of the lastball. The result is, that although the difference in time betweenthe falling of the various balls must be exceedingly minute, it issuch as practically to prevent all concussion. Lastly, the valvesconstructed on this plan are very easily repaired. It is only neces-sary to keep a few spare balls ready, to be inserted in the place ofany that may become damaged; and the old ones, melted andrecast in a mould kept for that purpose, are again as good as new.

Where it is proposed to work with a high pressure of steam, cutoff so as to allow of a considerable expansion, the beam engine isto be preferred to the direct-action engine; the latter, as a rule,