May 23, 1895] 



NA TURE 



91 



certain conditions must be obsen'ed to secure economy in ex- 

 pansive working. Unfortunately, more generally the conclusion 

 was drawn that the Cornish results were not to he trusted, and 

 that expansion was not economical, and Isherwood's own lan- 

 guage lent authority to the least accurate view of his results. To 

 obtain greater insight into the true action in the cylinder, and to 

 find a reconciliation of the Cornish and American tests, experi- 

 ments of a much more refined character were wanted and insight 

 due to wider scientific knowledge. 



The Physkal Properlies of Steam. — Rcgiiault. 



No useful progress could be made with a theory of the steam 

 engine, no accurate reduction even could be made of the results 

 of engine tests without exact determinations of the relations of 

 pressure, temperature, volume, latent heat and liquid heat of 

 steam. It was fortunate, therefore, that about 1840 M. Kegnault 

 obtaine<l the means from the French (iovi;rnnient to make a 

 series of researches on the physical properties of steam with 

 splendid instrumental appliances. He wisely carried out his 

 determinations over a very wide range of conditions, and spared 

 no labour or trouble in attaining accuracy. Kegnault's results 

 were of the greatest importance as a foundation for accurate study 

 of the steam engine. 



The Foitiidation of Thcniiodyiiaiiiits. — Carnot and fotile. 



The next important step was the discovery of the equivalence 

 of heat and work. Joule's investigations began with an attempt 

 to improve Sturgeon's magnetic engine. He was so led to 

 consider motive power problems from the engineer's standpoint, 

 as a question of duty, or of something obtained for something 

 expended. He ascertained the amount of electric current pro- 

 duced by the chemical combustion of a given amount of zinc, 

 and comjiaring his results with those obtained in good steam 

 engines, he concluded that, making the largest allowance for 

 possible imperfections of his magneto engine, it was never likely 

 to be a rival in economy to the steam engine. That was a negative 

 but a useful result. It closed one direction of useless endeavour 

 only too likely to attract the inventor. 



(jnc of the effects of electric action which Joule noticed was 

 the heating of his conductors, and it was to the measurement of 

 this heating eflect he next addressed himself. The heat developed 

 in the conductor by the electric action due to elements com- 

 bining in the galvanic cell was found to be identical with that 

 which would be generated by the direct combustion of the same 

 elements. Finally, he came to consider the relation between 

 the mechanical work expended in driving a magneto electric 

 machine, and the heat developed in the external circuit ol the 

 machine. He concluded that for 838 foot lbs. expended a pound 

 degree of heat was generated. Later experiments corrected 

 this value, but the discovery of the equivalence of heat and work 

 was made. 



.\s early as 1824, twenty years before Joule's discovery, .Sadi 

 Carnot, in a remarkalile pamphlet on the " Motive Power of 

 Heat," demonstrated the fundamental princijile that the amount 

 of work obtainable from any given quantity of heal cannot exceed 

 a quantity jiroportional to the fall of temperature. Unfortunately 

 adopting, though with hesitation, the view held in his time that 

 heat is material and iTidestructible as heat, he coupled with his 

 true jirinciple the false corollary that all the heat entering an 

 engine is discharged in the condenser. Joule, in 1845, found 

 this principle of Carnot, and looking to the corollary as es- 

 sential, supposed the principle itself to he false. He failed to 

 perceive that Carnot's principle was the essential sui^jilement 

 to his own di.scovery, and that it showed why the ajiparenl 

 efficiency of the steam engine is so low. It took six years 

 before Joule's and Carnot's principles w'ere reconciled, and for 

 three of them even Lord Kelvin refused to accejit Joule's 

 discover}', because it apparently conflicted \\ith the principle of 

 Carnot. 



J'hc Founders of I he Rational Theory, — Kan/cine, C/aiisins, 

 Zeuner. 



The impetus given to the study of thermodynamics by the 

 discovery of Jovde, and the perception of the fundamental im- 

 portance of Carnot's theorem, was enormous. Heat problems could 

 now be brought out of the region of mere empirical solutions, and 

 treated from the rational standpoint of an exact science, and the 

 steam engine, as the mo.st important examiile of heat transforma- 

 tion, attracted at once the attention of scientific men of command- 

 ing intellectual ability. In a very few years Kankine and 



NO. 1334, VOL. 52] 



Clausius had built up a strictly rational mathematical theory of the, 

 steam engine, and, a little later, Zeuner carried further the analysis 

 of some of the more subordinate details. The iheor)* with one 

 exception, to Ije referred to presently, took account of all the actual 

 conditions under which steam is used, and furnished exact rules 

 for the relation of steam expen<led and work done for all arrange- 

 ments of the actual steam engine practically adopted. 



It was just at this lime that the exjierimenls of Isherwood 

 were published, and a comparison of experimental results and 

 theoretical calculations showed directly a very large discrepancy. 

 The steam consumption in some trials was 30, 40 or 50 per cent* 

 more than it should have been in the assigned conditions of work- 

 ing according to the rational theory. Some action of quite 

 governing importance had obviously been neglected in the 

 theoretical analysis. 



The Experimental Theory. — Him and the .-ibatian .School. 



.\ year or two before Isherwood began his experiments, an 

 Alsatian engineer, M. Him, had discovered and measured 

 cylinder condensation. 



Joule's discovery attracted Hirn's allontion, and he set to 

 work in 1854 to verify, by an exact engine lest, whether the 

 difference between the heat received by an engine and discarded 

 in the condenser was the equivalent of ihe work done. His two 

 most important memoirs relating to the steam engine, are a 

 memoir on the utility of steam-jackets in 1855-6, and another 

 on the use of super-heated steam in 1S57. In these researches, 

 he devised a method of accurate engine tests, involving the 

 measurement of all the quantities of heat received by or rejected, 

 from the engine, which, with hartUy any change at all, is the 

 method of accurate engine testing adopted ever since. Under,, 

 his influence and direction, engine tests were carried out in, 

 Alsace for many years, and the results exactly analysed. It may. 

 be recalled that the admirable series of engine tests, the first; 

 tests in which the heat quantities were accurately measured in. 

 this country, which were made by Mr. Mair Kumley, and 

 described in three papers on " Inde])endent Engine Tests" in the 

 Proceedings oi this Society in 1882, 1885, and 1886. were trials 

 carried out strictly in accordance with Hirn's methods. 



.\s with Lord Kelvin, so with Him. It was the recognitioa 

 of an apparent conflict of Joule's di.scovery with Carnot's law 

 which first attracted his attention. It was the attempt toi 

 determine whether part of the heat svqiplied to an engine dis- 

 appeared as work which determined the form of his trials. His 

 experiments of 1S54 showed that " heat in a steam motor is not 

 only dispersed, but actually disaiipears, and the power obtained is 

 exactly proportional to the heat which ilisappears as heat to re- 

 appear as motive power." Some rather later and more careful 

 experiments enabled him to verify Joule's equivalent by the 

 actual results of a large engine test to an accuracy of about one 

 jier cent. 



The discovery of initial condensation, and the proof of the 

 powerful action of a small amount of heal Iransniilled from the 

 jacket, both pointed to the conduclivily .)f the cylinder wall as 

 the cause of the large waste of sleam which the constructors of 

 ihe rational theory had neglected. The cylinder is cooled 

 during expansion, and still more during exhaust by an action 

 analogous to internal radiation to the condenser. Before any 

 work can be done in the next stroke, the wall has to be reheated 

 by condensing fresh sleam. The extreme f^icility with which 

 steam yields or al.islracls steam by condensing and evajiorating, 

 accounted for the rapidity of ihe action. The m.agnitude of the 

 condensation increases with the range of temperature to which 

 the cylinder wall is subjected. It is larger in condensing than 

 in non-condensing engines, anil larger with high ratios of 

 expansion. 



Some lime ago I ventured to say thai there was no trust- 

 wiirthy engine test which showed that the consumption of steam 

 with a jacket is greater than without the jacket. I believe that 

 is .still true, but undoubledly the economy due to the jacket 

 varies in ditTerent cases from 30 per cent, to very nearly zero. 

 Roughly, the jacket is more useful with small engines than with 

 large ; with slow engines than with f;\sl engines ; but all this" 

 amounts to little more than saying that Ihe jacket is most 

 useful in ihose cases where the initial condcn.salion is largest. 

 Just in proportion as the engine, whatever its type, is of 

 the highest class and most scientific design, the jacket is less 

 useful. 



The jacket reduces, but it does not prevent initial conden^ition. 

 Ilirn looked for some more powerful way of healing Ihc 



