504 



NATURE 



[September 19, 1895 



'.inutras records of anemometers for long periods are required for 



rniining this pressure. The force of the wind also, and the 



• in. duration, and perio<l of occurrence of severe g.iles, are 



I'.ant to the maritime engineer for estimating the etTect of 



A.ues in any special locality, for determining the quarter 



.: _:.. which shelter is needed, and for ascertaining the seasons 



most suitable for the execution of harbour works, the repair of 



damages, and the carrying out of foundations of lighthouses and 



Iieacons on exposed rixks. The harbour engineer must, indeed, 



of necessity be somewhat of a meteorologist, for the changes in 



the wind and weather, the oscillations of the Ktrometer, and 



the signs of an ap|)roaching storm are indications to him of 



approaching danger to hus works, which he has to guard against ; 



for the sea is an insidious enemy which soon discovers any weak 



spot, and may in a few hours destroy the woik of months. 



Continuous records of rainfall, as collected regularly by Mr. 

 Symons from numerous stations in the United Kingdom, are 

 extremely valuable to engineers for calculating the probable 

 average yield of water from a given catchment area, the greatest 

 and least dischai^es of a river or stream, the size of drainage 

 channel nee<led to secure a low-lying area from floods, and the 

 amount of water available for storage or irrigation in a hot, arid 

 district. The loss of water by evajwration at difterent ])eriods 

 of the year, and under different conditions of soil and climate, 

 the effect of jiercolation in reducing evaporation, and the 

 influence of forests and vegetation in increasing the available 

 rainfall, while ei^ualising the flow of streams, are subjects of 

 equal interest to hydraulic engineers and meteorologists. 



Countries [leriodically visited by hurricanes, cyclones, or 

 earthquakes, necessitate special precautions, and special designs 

 for structures ; and every additional information .as to the force 

 and extent of these visitations of nature is of value in enabling 

 engineers to provide more effectually against their ravages. 



fifttcjils tonfcmd hy Eiii^int'trs upon Pure Scit'ihc. — Engineer- 

 ing is generally concerned in the application of the researches of 

 science for the benefit of mankind, and not in the extension of 

 the domain of |nire science, which necessitates greater concen- 

 tration of attention and study than the engineer in practice is 

 able to devote to it. Engineers, however, though never able to 

 repay the ever-increasing debt of gratitude which they owe to 

 jast and ))resent investigators of science, except in rendering 

 thc>e abstract researches of practical utility, have, nevertheless, 

 lieen able incidentally to promote the progress of science. Thus 

 mechanical science, by the construction of calculating machines, 

 the planimeter, integrating machines, the tide-predictor and 

 tidal harmonic analyser of Lord Kelvin, the self-registering tide- 

 gauge, and various other instruments, has lightened the lalx>urs 

 • if mathematicians ; whilst excavations for works, and borings 

 have assisted the investigations of geologists. The mechanical 

 genius of Lord Kosse led mainly to the success of the gigantic 

 telescope, which has revealed so many secrets of the heavens ; 

 and the rajjidily of locomotion, due to the lalxjurs of engineers, 

 has greatly facilitated astronomical ribservations and ])hysical 

 discoveries, Ixisides promoting the concourse of scientific men 

 and the diflusion of knowledge. Electrical engineering, more- 

 over, is so closely allied to electrical physics that the develo])- 

 mcnt of the one necessarily promotes the progress of the other. 

 The observations also conducted by hydraulic and maritinie 

 engineer^ in the course of their practice aid in extending the 

 statistics u|«n which the science of meteorology is based. 



F.ii.;iiiecring as an ExfvrimcHtat Science. — Engineering, so 



' it is iKused on mathematics, is an exact science, and the 



due to given loads on a structure can be accurately 



' : but the strength of the materials em|)loyed h.is to 



• (I before any structure can be properly designed. 



. the resistance of materials to tension, compression, 



. has to l)c tested, and their limit of elasticity and 



ight determined. Thus, previously to the constnic- 



tiua, by Kolicrl .Stephenson, of the JSrilannia Tubular Bridge, 



the f\m wriiiighl-iron girder bridge of large span erected, 



:ienlson various forms of wrrjugiit iron were 



It eminent mathematician and mechanician 



. :. ....... -'11, who had previously indicated the projX-'r 



theoretical form for cast-iron girders, and to whom the success 



of the fjridge across the Menai Straits was in great measure due 



("The Britannia and Conway Tubular Bridges," Edwin Clark, 



vol. i. p. 8?>. lie'iides the numerous tests always now made of 



n|iloycd during the progress of any large 



railway bridges are also snbjec(e<l to severe 



I>eing o|(cncd for public traffic, by uhich the 



NO. I 35 I, VOL. 52] 



safety of the structures and their rigidity, as measured by the 

 amount of deflection, are ascertained, serving as a guide for 

 subsequent designs. 



Numberless experiments have been made on the flow of water 

 in open channels, over weirs, through orifices, and along pipes ; 

 and the influences of the nature of the bed, the slope, depth, 

 and size of channel, have been investigated by various 

 hydraulicians. Mr. Thomas Stevenson measured the force of 

 waves at some places on the Scotch coast ("The Design and 

 Construction of Harbours," Thomas Stevenson, 3rd ed. \i\->. 

 52-56) : I'rof Osborne Reynolds has examined the laws of 

 tidal flow in a model of the inner estuary of the Mersey, and in 

 specially shaped experimental models (" British .Association 

 Reports" for 1S89, 1S90, and 1891); and I have found it 

 |X)ssible, in small working models of the Mersey and Seine, 

 not merely to re|)roduce the configuration of the bed of the 

 estuary out to sea, but also to oliserve the eflects of difterent 

 forms of training works in modifying sandy estuaries.' 

 Mr. William I'roude, after his retirement from active practice, 

 devoted his abilities to experiments on the motion and resistance 

 of ships in water, which have proved of inestimable value to 

 the naval architect, and which formed the subject of his 

 presidential address to this Section in 1875. 



Electrical engineering is specially adapted for experimental 

 investigation ; and, in this branch, theory and practice are so 

 closely allied that some of the most eminent exi>onents of the 

 theor)' of the subject, such as Lord Kelvin and Dr. Ilopkinson, 

 have developed their theories into practical results. In most 

 other branches, the investigator is generally distinct from the 

 engineer in large practice ; but it may be safely said that an 

 able investigator and generaliser in engineering science, as, for 

 instance, the late Prof. Rankine, accomplishes work of more 

 value to the profession at large than the jiractical engineer, 

 'who, in the world's estimation, appears the more successful 

 man. 



Every branch of engineering science is more or less cai>able of 

 being advanced by experimental investigations ; and when it is 

 borne in mind that the force of waves, the ebb and flow of tides 

 in rivers, the influences of training works in estuaries, and the 

 motion of ships at sea have been subjected to experimental 

 research, it appears impossible to assign a limit to the range 

 of experiments .as a means of extending engineering knowledge. 

 Trobleins of considerable interest, which can only be solved by 

 experiments or by contprehensive generalisations from a lunnber 

 of examples, must frequently present themselves to engineers in 

 the course of their practice, as they have to myself ; and 

 engineers would render a great service to the profession if they 

 would follow up the lines of investigation thus suggested to them, 

 in the true spirit of scientific inquiry. 



r'ailurcs of IVi'i h due lo NcglccI of Scicniific Considerations. — 

 Before the amount and distribution of the stresses in structures 

 were thoroughly iniderslood, a disposition was naturally evinced 

 to err on the side of excessive strength ; and the materials in the 

 various parts of the structure were not suitably proportioned to 

 the load to 1«; borne, resulting in a waste of materials and too 

 great an expenditure on the works. Thus some of the early 

 high m.isonry reservoir dams in Spain exhibit an excessive tliick- 

 ness towards the top, imposing an unnecessary load on ihe 

 foundations ; and in many of the earlier iron girder bridges 

 more material \\.as employed than was required for stability, and 

 it w.as not pmiK-rly distributed. Boldness engendered by 

 increased experience, and dictated by motives of economy, has 

 (ended to make the engineers of the present day pursvie an 

 op|K)site course ; and, under these circumstances, tlie correct 

 calculation of the strains, the exact strength of the materials, 

 and a strict appreciation of the physical laws aftecting the 

 designs become of the utmost im|M>rtance. 



The failures of many bridges may be explained by errors in 

 design, defects in construction, or by economy carried beyond 

 Ihe limits of safety in inishing forward railways in undeveloped 

 countries : but other f;iilures are attributable to a disregard or 

 underestimation of the influence of physical causes. Tluis ihe 

 Tay Bridge disiister, in 1.S79, was due to underestimating the 

 amount and effect of the wind-pressure in an exposed situation, 

 where it acted Hith a considerable leverage, owing to the height 

 of the bridge, and was inadequately provided against by the 



1 J'rtKffiiings of Ific Roy.il Society, vol. xlv. pp. 504-524, and pKitc^ i-i \ 

 vol. xlvii, p. 14:1 ; .niid " Aln^liorHlioii dc la P.irlic M.iritiinc des Hcuvt^s, y 

 conipri<i lcuf% KinlM.tin;luircs," I.. I'. Vcrnoii-H.'ircourt, l^iiris Inlund Naviga- 

 lion Congrcw,, tSyj, pp. 37-29, and 33, ^y, and plalc 3. 



\ 



