{86 



NATURE 



[August i6, 1894 



natural matters, and in too many cases the results of their labours 



:£E^^e;t^.f;:c:ll^^.S:r^i^AH 



7^ than ome solitary megatherium «ould be wbo hav.ng 

 o^ived Si his contemporaries. «as able in after ages to look 

 down upon his pigmy and short-lived successors Mere length 

 STsuch a^case may be a mere accident, and .s "0. -ts If a 

 proof of fitness. We have it thrown at us every now and then 

 Kat ou engines nowadays do not last l,ke the old ones as f 

 the mere existence of a very old machine were a proof of its 

 l^ues is certainly " proof of the excellence of ..s con- 

 Imiction-or, as one may say, of its const.tut.on-and perhaps 

 al™ of the very small amount of work it has done m proportion 

 to its life and its dimensions. 



It is sometimes, I am afraid, rather hum. hating to have to 

 remember that, to a very great extent, the question of the finest 

 so far as it affects us, is a financial one. In manufactur ng 

 processes efficiency and economy tend to ^""'"l >^ ""^^ 

 they lead to decreased cost of production. In structures or 

 other large permanent works those types tend to perpetuate 

 "hemselves which require the least material-that is, in which 

 the material used is disposed to the best advantage-and in 

 which the outlay on labour is also smallest, ^^^"'"•"f • °f ^°";.^^; 

 equal fitness in other respects. There is, no doub »'/'"«"' 

 aCdency to dispute this' altogether, and to treat all reducuons 

 in cost of labour as disadvantageous, unless, indeed, he labour 

 be very highly skilled, in which case its remuneration must 

 necess7rily1>e brought down lor the sake of equa ity I imagine 

 Xs tendency will last exactly as long as the j^»'>h ". "" ge 

 some other people to pay the increased cost, and will therea ter 

 deTen^ine itself somewhat suddenly. It can no more stand n 

 the ™y of natural progress in engineering matters than could 

 thrso'mewhat sim'ila? outcry against, the -trofct'O" °f 

 machinery into manufactures two generations ago. It «°"''l ^= 

 Z wise to paint a generation of cats green, in the hope of com- 

 oellinc natural selection lo woik along new lines. 

 '^ I thik we may fairly assume, therefore, that efficiency and 

 economy are both legitimate criteria as to ultimate fitness, an 1 

 t'll remain so. Moreover, they are both matters in which 

 measurements can be made, and as •V?"^^^:'^" other 

 be guided by such measurements. But there are other 

 charic eristics not directly measurable, by which we can in 

 some degree fcrm an opinion as to the ultimate fitness of things or 



^' One ?et of considerations which has great critical importance 

 is summed up in the word .m//:V,7v. This does not mean few- 

 nessTparts^ Rculeaux showed long ago that with machines 

 there waj in every case a practical minimum number of pans, 

 any rercion below which was accompanied by serious pracUcal 

 drawbacks Nor is real simplicity incompatible with cons.der- 

 aWe appa nt complexity. The purpose of machines is becoming 

 conlinually more complex, and simplicity must not be looked at 

 M absolute, but only in its relation lo a particular purpose. 

 The e a "many veiy complex-looking pieces of apparatus in 

 Existence which^ork acluklly so directly along each of their 

 "any branch lines as to be in reality simple. I Relieve it almost 

 always happens that the first attempt to carry out by » "I'ch ne 

 anew purpose is a very complicated one. It is only by the 

 closest To-l-le examinaTion of the problem the Retting a its 

 very essence, that the machine can be simplified, and this is a 

 bt^and not "an early stage of design. If ^l^"*^?"'" P'°^ ^f, 

 is leally only soluble by exceedingly complicated apparatus, .1 

 genefally becomes a que.stion whether the solution is worth 

 Taving. ^ There is no imposMbility in making .=J -"j-^hine tha 

 will do anvthing. But the very simplest possible fo m of ap 

 paratus Xh would wash our hands for us in a suitable manner 

 f" . ,.,,..„ very complicated that for many years to come at 

 1. ration will be performed by manual labour. 



. ly allied lo simplicity is what 1 may call directness. 

 In nearly all mechanical processes certain '""'•!"""''"°"' "^ 

 unavoidable. In many mechanical processes, as I have recently 

 ) ,n to mention, a very large number of transformat ons 



, . [.radically unavoidable. I myself cannot help 

 • - I ably one of the most distinct signs of fitness 



rr of transformations, the bringing of the final 

 ngcs as close together as possible, and cutting 



out altogether the apparently worthless middle processes But 

 any generalisation of this kind must be very cautiously handled ; 

 these apparently useless processes are no dcubt in certain cases 

 as SSle as is the much abused middle-man in matters 



"TnTcritical view of any case where similar results "e aimed 

 at bv hand-work and by mechanical means, it is important to 

 recognise that the similarity of result should very seldom become 

 identity. In the first machine to do anything mechanically 

 wMchLs before been done by hand, ^^eerror is often made of 

 trying to imitate the hand- work rigorously. The f'^^' se^""! 

 machines were, I believe, made to stitch in the same w^ay as a 

 Team "ress It was not until a form of stitch suitable for a 

 macTiine. although unsuitable for hand, -- <>--e^„'^^';' 'J^ 

 sewing machine proved successful as a P"<=' "\ T'^*""; . \° 

 another but analogous line too you may remember th.at the first 

 railway carriages tere practically stagecoaches put upon trucks 

 }rom which .the present carriages have only very slowly been 



"'Thf critic has also to remember that very often the attain^ 

 ment of some very unimportant point, or point of "hich he 

 importance has been greatly exaggerated, is made the reason 

 iTechanicallv for very great complication. The question of pro- 

 portn co2s in her^'e'again, and it has to be conf-ed in any 

 particular case whether the academically perfect machine, «h eft 

 is also extremely complicated, is not '"f"'°\'° . '"l^"'"?'' 

 equally good machine which has been constructed in a practic 

 alilp shaoe — it almost always is so. 



"' Vave'e^deavoured in my'remarks to infi"'V\^^' ^'^C" 

 to me to be the attitude of the engineer towards » >''y '^'f* 

 portion of the work which comes into his hands. In order to 

 deal with the work it is necessary for him first of all to have 

 a certan definite knowledge of *• things," that is lo say bo ho 

 fhe various subjects which foim part of the curricula of id 

 echnical schools, and of the further matters which form as it 

 were hTs profetsional alphabet. These last he 'earns not from 

 books or lectures as a student, but by example •'>n<l .»"«"'? '^^ 

 does an a fis . Of this part of his training I have said nothing ; 

 ft has been l^erhaps sufliciently talked about of ate years, and 

 here is little to s^ay which 1 could h.tve made interesting to a 

 ceneral gathering like this. I cannot leave it a together, however, 

 witbout^deaUnB with one matter. Kxceptional men are all- 

 Tound mathematicians or physicists, ^'»l "'"^ -"^'^^ X 

 are both : but for ordinary folk the study of one side ol ma ne- 

 mitics o; of a single b.anch of physics is the woik of a life- 

 ,Tme The engineer is bound lo know his own profession, by 

 rpolhesis and it is in itself no small matter. ^ et in addition i 

 h^Ct know some mathematics, some physics some chem stn^. 

 even also some geology, if he is to take an ■ high r^n"^ '" ' • 

 It is therefore, surely in the very nature of things ■nipossib e , 

 tha he should be a great mathematician or a great physicist, 

 or should devote af much study to those "los. ascinating 

 sciences as if they themselves were the work of his life 1 here 

 fore I beseech my friends of Section A to do what ihey can to 

 Modify their natural attitude of superiority--even of contempt 

 -towards us, especially when we arc students. The young 

 ennineei-ll peak as a member of the great ma.tority of he 

 ordinary kind^would probably never have chosen his profes- 

 sion if he had had special aptitude for m.ithemat.cal work. 

 Having chosen il, he 'has to look at -'^-rreir'T '"'an'o 

 tool a means to an end, not an end in itself. '"""«' 

 my elf see thai this point of view is one -I'^'^I'" "' '° 

 Te parent of all the sciences, and 1 am confirmed by the 

 knowledge that one or two of the greatest n,all.emat.cian» 

 ;n ?W rountrv are of the same opinion and have the courage 

 oL on ."-will infinitely' beneficial icsul.s to the 

 'young men they have to .leal with. But I know la to ma h^ 

 ^naticians in general-the physicists »'<= .""f ^i^''"^;;^;''^ "^Z 

 name of engineering student is o<''''"»;.'"'''^f ?^i, f/^Lct." 

 who wilfully refuses to make m.ithematics his first su jeci, 



r d :ho,l/erefore dese.ves neither --•^^"''-..^"^..rerto 

 to whom it is privilege .sulTicient that he ='»'°"; ''' ^ ""j for 

 pick up such crumbs as he can digest f.om »'=',';''f,'\'^ ;;;,,„,. 

 his betters. I humbly Jirolest that we deseive better "ealmen'. 

 Hi no doubt a great'iliisfortune to us that we cannot afford to 

 spen.l our training-time preparing for <=»^"""" ""''; "" ,e?in 

 we have been compelled to choose for ;<"; f"'7 •\\' f U" 

 which mathematics plays only a ^ccon.lay part. It - 

 further misfortune that we have to solve '"enl) reaM.ve p 

 blems, each demanding a real live answer, for every single o 



NO. 1294, VOL. 50] 



