26 ANNUAL OF SCIENTIFIC DISCOVERY. 



of the Naval Engineer Corps, to determine certain questions in refer- 

 ence to the economy of steam expansion. Previous experiments 

 made by the chief officer of the Board had induced him to assert the 

 fallacy of the commonly received doctrine of economy in expansion, 

 and these observations were undertaken to pursue the investigation 

 on a more perfect engine, and with greater care. A report of the 

 results has been published by the Navy Department, from which the 

 following facts are derived : 



The vessel selected to test the relative merits of expansive and 

 non-expansive steam in cylinders was the Michigan, a government 

 paddle-wheel steamer lying at Erie, Pa. The larboard engine only 

 was used, and it was employed in exerting its power to paddle the 

 water aft while secured at the dock. Each experiment lasted seventy- 

 two consecutive hours, during which the engine was neither stopped 

 nor slowed down, nor in any way changed in condition. It was always 

 operated several hours, so as to get the steam to the same pressure, 

 the fires in proper order, and all things adjusted correctly, before each 

 experiment was actually commenced. The water in the boiler was 

 gauged, and the quantity fed in was accurately measured. Every 

 pound of coal fed into the furnaces was carefully weighed ; indicator 

 cards were taken, and everything arranged to insure accuracy. The 

 results of seven experiments, cutting off at ^, ^ -|, -j 3 ^-, -|, ^, -j^ths of 

 the stroke, are given in a tabulated form. Five of these were perform- 

 ed with Ormsby bituminous coal, and the other two with anthracite and 

 Brookfield coal. The pressure in the boilers above the atmosphere 

 was 195 Ibs. the lowest, 22 Ibs. the highest. The quantities of water 

 consumed were 39.942 Ibs. per total horse power, cutting off at 1^- 

 stroke; 30.881, at T 7 -g-; 29.416, at ^; 30.592, at T 3 g-; 29.841, at J; 30.715, 

 at 1-; 32.044, at -%. These are important items, demanding careful 

 scrutiny. 



The water fed into the boiler was carefully measured in a tank, 

 and it was found that just in proportion as expansion was extended, 

 there was a proportionally greater loss of steam in the cylinder by 

 condensation a great deal more steam flowed into the cylinder 

 than was accounted for by the indicator. Thus, cutting off at l-| of 

 the stroke, the loss was 2.91 ; at T 7 ^, 6.60 ; at -f, 18.14 ; at T 3 , 33^7 ; 

 at l, 30.84 ; at l, 33.66, and at -%, not less than 37.16 per cent, not 

 accounted for. In short, the conclusions from these experiments are, 

 that so much condensation takes place in the cylinder by the cooling 

 action of expansion, that no economy results from using highly ex- 

 panded steam. It is true that somewhat more power is developed in 

 the same cylinder by expansion, but by using smaller cylinders, pro- 

 portioned to the power required without expansion, the economy is 

 on the side of non-expansion, both with respect to fuel and the cost 

 of machinery. These conclusions, it must be evident, are of a very 

 radical and revolutionary character, inasmuch as they affect principles 

 which have been accepted in practice from a very early period in the 

 history of the steam engine applied to actual work. They differ from 

 the whole tenor of experimental observations and theoretical deduc- 

 tions, and, if accepted by the profession, would modify at once our 

 proportions of working parts, and our applications of power. But it 

 should be also stated, that these conclusions are not universally ac- 



