94 



NATURE 



[November i6, 19 



measure of constriiclive value, though it certainly is a 

 valuable indicatiun. Furthcj-, the clastic limit in com- 

 prussioii is srldoni ohsfrvf-d, and iho clastic limit in shear 

 is «'xp<"rim(!niiilly aiinosi unknown. Yet resistance to shnar 

 is probably for ductile materials the nuist important 

 element of constructional value. It would be interest inj^ 

 to have the elastic limit in torbion ©f thin tubes accurately 

 determined. 



Hence, whether we take the breaking stress, the yield 

 stress, or the primitive elastic limit, we have not found 

 a satisfactory rational basis for measurinjj the constructive 

 strength of a material. There is not a fixed ratio between 

 the greatest safe working stress and either of these. But 

 1 do not wish to be understood as depreciating the value 

 of the determination of breaking stress, yield-point, and 

 elastic limit in testing materials. 



Valuable as are the data of breaking strength, yield- 

 point, and elastic limit determined in a tension test in 

 influencing the judgment of the engineer, they do not 

 furnish any purely rational rule for fixing the working 

 stress for designers. Kxperience in <similnr ca^e^; must 

 always be the ultimate appeal. 



Compound Stress. 

 So far, all that has been said delates to cases of simple 

 tension, thrust, or shear. But in a very large number of 

 cases two of these stresses are combined. In such cases 

 the theory of elasticity furnishes the principal stresses and 

 the principal strains ; but the question. What is the 

 criterion of safety? involves another of the unsolved 

 problems of engineering. Two schools arose amongst 

 theorists, one holding that the maximum stress, the other 

 that the maximum strain, determined fracture. For 

 simple stresses, either assumption leads to the same result. 

 It is not so with compound stress. Very early Coulomb 

 suggested that fractures would be determined by shear, as 

 is the case with cast iron under compression. The subject 

 is now under investigation experimentally, and, speaking 

 generally, it seems that for ductile materials the material 

 gives way at some limit of shearing stress, while for 

 brittle material in tension the principal stress is the best 

 criterion of strength. 



Comparative Structural Value of Materials. 



It is only when comparing two different materials that 

 an engineer needs to consider relative constructive value, 

 and then he bases his judgment on all the properties of the 

 material. Ordinarily, the safe working stress is fixed by 

 experience in similar cases, theory helping mainly in ox- 

 plaining why in different cases different stresses are suit- 

 able. The history of engineering design is a history of 

 experiments in construction, in which by trial and error 

 right proportions have been found ; and in the settlement 

 the different relative importance of cost and weight enter 

 into the question, as well as considerations of mechanical 

 strength. The fact that scientific data and rational, or 

 semi-rational, formulas are conveniently used should not 

 be allowed to conceal the fact that an empirical element 

 always enters into the solution. 



New designing is really a process of comparison, in 

 which the engineer extends experience in known cases to 

 new conditions ; and the problem is in what proportion 

 dimensions must be varied to allow for differences of size, 

 of material, of loading, of speed, or of form. 



Most practical problems in designing for strength 

 reduce themselves, if traced to their foundation, to applica- 

 tions of the law of similarity. At a time when steam 

 engines all worked at about the same steam pressure, it 

 was broadly stated that the drawings of an engine of any 

 size could be used for constructing engines of any other 

 size by merely altering the scale of the drawing. So-called 

 mechanical instinct is, no doubt, really reasoning based on 

 the law of similarity. 



In the case of machines, another consideration enters. 

 The weight stresses are not usually important, but the 

 inertia stresses due to variation of velocity or direction of 

 motion are very important. The stresses in a fly-wheel 

 rim due to radial acceleration are proportional to the 

 square of the rim velocity, whatever the cross-section. 

 Hence fly-wheels are equally safe in this respect if the 

 rim velocity is the same, or if the rotations per minute 

 are inversely as the diameters. 



NO. 2194, VOL. 88] 



Engineering Education. 



I turn to another subject, which seems to be , _ 

 in a year in which the institution has held a coi 

 on the education and training of engineers. An inij. 

 change is going on in all types of education. In all it 

 being recognised that they should Ijc such as to afford 

 training for the duties of life. The term "engin*-: " 

 used in such wide and loose senses that it is m 

 to be explicit as to one's point of view. There arc .«. 

 tions associated with engineering work for which t, 

 special training is absolutely required Ix-yond that of il 

 factory or commercial office, or for which any other traini; 

 required is different and more limited in scope than th 

 of the professional engineer or director of engii. 

 works. But the conference considered the case <>l 

 who aim at becoming ultimately professional en;, 

 that is, men who advise, design, and direct the ex' 

 of works of civil engineering in the wide sense in 

 that term is used in this institution. If the course > 

 training of such men is provided for, the less compl- 

 training of men for lower grades of service is not 

 to be wanting. But I do not think that we dist; 

 sufficiently at present the different type of education i 

 quired for leaders and subordinates, the heads and tl 

 hands. 



It must be recognised tliat the professional or consul 

 ing engineer is not solely concerned with technir. 

 problems, though as to these he must be an expert. Bi 

 he is more and more concerned with economic, legal, ai 

 commercial problems of much intricacy, and must be pr- 

 pared to meet men of affairs and of liberal education v. 

 an equal footing. 



Now the earlier great engineers in this country had r 

 formal technical training; indeed, very little gerteral edu( ; 

 tion either. Brindley, George Stephenson, Fairbairn, h;; 

 only parish-school education; Smeaton, Telford, and Wa 

 only grammar-school education. They picked up even th- 

 practical experience gradually and casually. They differ^ 

 in this from their contemporaries in France. Perronei, 

 Gauthey, Rondelet, and Navier, for instance, were of a 

 more academically educated type. 



It is, no doubt, due in part to such early conditions < 

 English engineering, in part to the narrow and unscie; 

 tific character of secondary and university education in th 

 country, that imtil recently the education and training i 

 engineers has been so unacademic and unsystematic. 



Preliminary Education. 



A very great obstacle to the progress of technical educ. 

 tion in this country has been the want of any definite air 

 — the unsuitable character and poor average quality of oi: 

 secondary education. Grant whatever can fairly be urg' 

 as to the honourable spirit, the pluck, and resourcefulm- 

 of boys from the better public schools ; grant also tl. 

 value of dead languages as a means of culture for tho- 

 who really master them, it must still be recognised th. 

 the average boy, to some extent, still leaves scho. 

 sadly unfitted for any form of higher education. Tlv 

 success of the clever and brilliant few, educated for un: 

 versity prizes, has obscured the poor quality of the intf ' 

 lectual equipment of the average many. 



Happily, if the state of things is not yet satisfactory, ii 

 secondary education is still unorganised, and the school- 

 master is only partially converted, still science has obtained 

 a footing, and scJiool-leaving and other examinations hav^ 

 secured that the average lad is better taught. 



It was partly a cause, partly a consequence, of defectiv 

 school education that the great universities have nt 

 addressed themselves, in a responsible and scientif 

 manner, to securing that candidates for admission ; 

 higher courses had reached a reasonable standard. Sti! 

 less were the engineering schoolsj with one or two excep 

 tions, able to insist that students were properly prepare 

 for technical instruction. It is only recently, and partly 

 under pressure from the institution, that a fairly satis- 

 factory entrance examination has been prescribed in som 

 English and colonial engineering schools. To admit lad- 

 with imperfect preliminary education to advanced instruc- 

 tion lowers the whole standard of work in the classes. 



An erroneous opinion has prevailed that technical instruc- 

 tion requires little or no literary and scientific education t' 



