744 REPORT — 1894. 



experiment is tbe best which ensures the greatest possible accuracy in every part 

 of the result. The business of the engineer, on the other hand, is to deal with 

 physical problems under conditions which he can only very partially control, and 

 the conditions are a part of his problem. He does not, for instance, experiment 

 with a steam engine so made that it can work with a Carnot cycle. It is in the 

 nature of the case that he must experiment with a much less perfect machine. In 

 burning fuel he does not use apparatus specially made to absorb the whole heat of 

 combustion, but in the nature of the case has to investigate the behaviour of appa- 

 ratus in which a very large part of that heat is unavoidably wasted. So one might 

 go on through an immense number of instances. Perhaps the whole matter may best 

 he summed up by saying that in a physical laboratory the conditions of each experi- 

 ment are under the control of the experimenter, and are subservient to the experi- 

 ment itself. In an engineering laboratory the conditions form part of the experiment. 

 However much more ditlicult or complicated they render it, they still unavoidably 

 form part of it — an experiment under any other conditions, or with those condition* 

 removed, would ipso facto he irrelevant. 



A critical training in matters mechanical is, however, only too similar to the 

 celebrated training of the Mississippi pilot which so nearly broke the heart of 

 Mr. Mark Twain. Whenever the whole matter seems to be completely mastered 

 from one point of view, it is only to find, with a little more experience, that from 

 another point of view everything looks different, and the whole critique has to be- 

 started afresh. Machines cannot be finally criticised — that is to say, they cannot 

 be pronounced good or bad — simply from results measurable in a laboratory. One 

 wishes to use steam plant, for instance, by which as little coal shall be burnt as 

 possible. Ikit clearly it would be worth while to w-aste a certain amount of coal 

 if a less economical machine would allow a larger saving in tbe cost of repairs. Or 

 it might he worth while to use a mnchine in which a certain amount of extra 

 power was obviously employed, if only by means of such a machine the cost of 

 attendance could be measurably reduced. In fact, what may be summed up in 

 the phrase the ' worth-whileness ' of economies, is in itself a matter on which a 

 whole paper might be written. Unfortunately, the latter points which I have 

 mentioned are just such as cannot easily be measured in laboratory work, or, 

 indeed, in any other way whatever, except by actually using the apparatus im 

 question. All that can be said is that a careful training in the critical measure- 

 ment of comparatively simple points fits a man more than anything else to gauge 

 accurately the importance of such other matters as I have mentioned. No doubt 

 there are many men in whom the critical faculty is insufficiently developed to- 

 allow them ever to be of use in these matters, but to those who are intellectually 

 capable of the ' higher criticism ' it must be, I think, of inestimable benefit to 

 have had a systematic training in the lower. 



Is there, then, any general standpoint from which mechanical criticism can he 

 directed ? Certain points are obvious, but probably the whole matter cannot 

 easily be gener.alised. A city has to be supplied with water; there are three 

 requisites : that the water should he of proper quality, of sufficient quantity, and 

 that it should be brought in at a reasonable cost. But in such a case the first two 

 are so enormously more important than the third, that the ideal is comparatively 

 simple (of course, this is quite a different thing from being simply reached). A 

 city has to be supplied with electric light : the essential conditions are similar. 

 But in this case there are so many qualities which are equally proper, and 

 there are so many different ways of bringing it in in sufficient quantity, that the 

 third point — namely, the cost — becomes especially important. A factory has to 

 be driven by steam power : the amount of power that is wanted can be produced 

 by so many different types of engine and boiler — all capable of approximately equal 

 economy, and all claiming equal freedom Irom breakdowns — that the choice is a 

 peculiarly difficult one from the critical point of view. 



It seems almost impossible that a criticism on any one basis could meet all the 

 three cases which I have supposed, unless that basis were that the thing supplied 

 should be the absolutely fittest, having regard to all the conditions of each case 

 and the relative importance of each condition. Possibly in all cases we could get 



