December i6, 1922] 



NA TORE 



S2: 



biological science. The fact must, however, be faced 

 that in all cases the observing instrument is a living 

 organ and is, therefore, in a perpetual state of change. 

 The rate of change is relatively slight in the most 

 favourable cases, but rapid and complex in the less 

 favourable. Physicists have been notoriously suc- 

 cessful in so reducing the physical complications of 

 experiments to a minimum that the problem nearly 

 approximates to a mathematical abstraction, and, 

 therefore, the highest degree of accuracy. Further 

 advance is to be sought bv greater attention to the 

 biological complexities in order that they, too, may 

 be subject to more complete control. 



A mass of evidence has of recent years accumulated 



to show that in peripheral vision two mechanisms 

 are simultaneously at work. Of these, one is chiefly 

 concerned with vision under low intensities of light — 

 what I have called scotopic vision. The end organ 

 of this mechanism is the rods of the retinal neuro- 

 epithelium. Photopic vision, or what may be called 

 daylight vision, is chiefly carried out by the cones. 

 The duplicity theory is so well established that it has 

 even found its way into the writings of the physicists. 

 The explanation and our knowledge of retinal adapta- 

 tion depends upon these physiological facts. Since 

 retinal adaptation plays a preponderant part in 

 simultaneous and successive contrast its importance 

 in photometry will be reaclilv realised. 



The Design of Railway Bridges. 



A SUBJECT of great importance to the general 

 1 •*■ public is the safety of the thousands of 

 bridges by means of which our railways cross roads, 

 rivers, and other railways. Probably it occurs to 

 few railway travellers to consider the complexity 

 of the design of each bridge they cross and the 

 organisation required to inspect, test, and maintain 

 every bridge in a condition suited not only to the 

 traffic for which it was originally designed, but also 

 to the increased weights and speeds which have 

 since been introduced. It is but natural that differ- 

 ences of opinion should arise between the railway 

 companies which have to pay for their erection and 

 maintenance, and the Board of Trade which has 

 to satisfv itself that they are safe. 



The Ministry of Transport has recently carried 

 out a series of tests on actual bridges, and has issued 

 a report containing suggestions which appear to 

 foreshadow regulations requiring railway bridges to 

 be heavier and therefore more expensive. This 

 report has naturally aroused great interest and caused 

 no small concern among the bridge engineers of the 

 leading railway companies. At the meeting of the 

 British Association at Hull, the Engineering Section 

 devoted a morning to a discussion of the problem. 

 Unfortunately no representative of the Ministry of 

 Transport took part in the discussion, but the railway 

 companies were well represented and the speakers 

 included the bridge engineers of the Great Western, 

 North Eastern, and Great Central companies. Taken 

 together the papers constitute a concise but fairly 

 complete review of the present situation. 



Mr. J. S. Wilson, who opened the discussion with 

 a general review of the questions involved, showed 

 that the difference of opinion between the companies 

 and the Board of Trade is nearly as old as the rail- 

 ways themselves. In 1849 Torksey Bridge across 

 the Trent was tested by the representative of the 

 railway commissioners preparatory to the opening 

 of the Retford and Lincoln line. The deflection of 

 \\ inches with four locomotives and tenders on the 

 centre of a span was considered excessive and per- 

 mission to open the line was refused. The bridge 

 had been designed by John Fowler who, with Sir 

 Benjamin Baker, was responsible later for the Forth 

 Bridge. He had followed the rules laid down by 

 Fairbairn, and he suggested to the commissioners 

 that some mistake had been made, but after further 

 tests the latter persisted in their view that the 

 stresses in the bridge were excessive. Finally, how- 

 ever, Fowler succeeded in convincing the commis- 

 sioners that the girders, being continuous over the 

 middle pier, were not stressed so highly as would 

 otherwise be the case, and on his offering to reduce 

 the weight of ballast on the bridge, the line was 

 finally opened after a delay of three or four months. 



NO. 2772, VOL. IIO] 



The subsequent history of the bridge is of interest. 

 The bridge is still there; for forty-six years it was un- 

 altered and carried all traffic satisfactorily ; in 1896 

 it was strengthened by the addition of a longitudinal 

 girder. These old iron bridges designed by Fairbairn 

 and Fowler, which have stood the test of seventy vears' 

 wear and tear and are still in good condition, are 

 powerful arguments in favour of the view that 

 bridges built on the same assumptions will be per- 

 fectly safe. 



There are many difficulties, however, in the cal- 

 culations and assumptions involved in the design, 

 in allowing for the effect of impact due to the fact 

 that the load is a live one, that is, not a stationary 

 load, and in allowing for the effect of sleepers, rails, 

 and ballast in strengthening the structure, distribut- 

 ing the load, and damping out the effects of impact. 

 It is here that some doubt arises as to whether the 

 intentions of the Ministry are correctly interpreted 

 by the railwav companies ; it is useless to specify 

 a factor of safety or a working stress unless one also 

 specifies how the stress or factor is to be calculated 

 or determined. The bridge designer may employ 

 what appears to be a low factor of safetv because 

 he knows that the actual stresses are less than 

 those calculated by the simple conventional methods 

 usually adopted and that his actual factor of safety 

 is consequentlv much greater. If a high factor of 

 safety is specified, then it is open to the engineer to 

 modify not his design but his methods of calculation 

 so as to take account of the various strengthening 

 factors usually neglected, and thus obtain a lower 

 calculated stress and a higher factor of safetv than 

 would be given by the usual semi-empirical method. 



One speaker in the discussion advocated making 

 full - scale tests on old bridges which were being 

 replaced ; these could be re-erected and thoroughly 

 tested, if necessary to destruction. All the speakers 

 deprecated the premature promulgation of rules 

 which would lead to heavier and therefore more 

 expensive bridges, but urged that present practice 

 should be followed until systematic research has 

 been carried out and far more knowledge of the 

 subject obtained than that on which the Ministrv of 

 Transport are proposing to act. 



In their attitude towards riveted structures of 

 iron and steel, engineers may be divided into pessimists 

 and optimists, and Mr. Wilson's experience showed 

 that the greatest optimists have been those most 

 closely associated with the maintenance or actual 

 construction, who would certainly be the first to 

 detect any indication that the bridges were showing 

 signs of weakness. 



An engineering student is always taught that the 

 stress produced by a live load is double that pro- 

 duced by the same load when steadily applied. This 



