G.— ENGINEERING. 147 



place, nor is there time, to consider in detail either the early or later 

 experiments, but only to refer to them to illustrate how rigidly controlled 

 experiments were essential to the development of metal bridges and 

 aeroplanes. 



There is still, however, a great deal to be done in connection with 

 structures, and the only hope of a satisfactory solution of many problems 

 is not simply by accumulating experience, but by combining experience, 

 experiment, and mathematical analysis in an endeavour to reduce the 

 whole to definitely co-ordinated knowledge. Nearly thirty years ago I was 

 privileged to assist with some experimental work on the deflection of 

 bridges produced by rapidly moving loads. They were incomplete, but 

 they definitely indicated that the assumptions often made by engineers 

 as to the effect of rapidly moving loads were not justified. In view of 

 the publication in the early part of this year of the report of the Bridge 

 Stress Committee, published by the Department of Scientific and Industrial 

 Research, of which perhaps it is only necessary to say that it is worthy 

 of its distinguished chairman. Sir Alfred Ewing, and his colleagues, it is 

 fitting to particularly refer to this subject of impact on bridges. 



As early as 1849, in the Report of the Royal Commission on the use of 

 wrought iron in railway bridges, the subject was discussed and Sir George 

 Stokes gave a mathematical solution for a load moving over a girder, 

 but this solution was of little value in practice. For eighty years the effect 

 of travelling loads has been treated in an empirical illogical manner and 

 a number of impact formulae, which the Bridge Stress Committee Report 

 shows have no valid foundation, have been rather blindly used by 

 engineers. The report shows very clearly that what is important is not 

 the fact that the load is appHed suddenly to the structure, but the possi- 

 biUty of synchronism or otherwise of the structure with the periodicity 

 of unbalanced forces due to the engine or to other causes. Experimental 

 data of the vibrations produced in the structure by the various types of 

 moving loads have been carefully obtained and methods of analyses 

 developed which have made it possible to determine with some degree 

 of precision the true impact factors to be applied. Thus a problem of 

 real difficulty, in connection with which doubt and uncertainty has 

 probably led to expenditure far greater than the necessities of the case 

 demand, comes within the possibilities of solution. The carefully conducted 

 work of the Committee spread over a few years has given more precise 

 guidance for future designers than a hundred years of practice. It may 

 quite reasonably be said that practice in the past has led to safety. That 

 certainly must always be the first consideration, but undue safety involving 

 the use of too much material or even too narrow a margin of safety by 

 using materials unsatisfactoril)'' cannot be said to be good engineering, 

 and engineers as well as the general public- — for whom engineers are 

 trustees — cannot be satisfied until all uncertainties that can be removed 

 by careful scientific inquiry give place to more accurate knowledge. The 

 only hope of completely solving this all-important problem, as well as the 

 equally important problem of the distribution of load through various 

 types of floors to the girders of both steel and reinforced concrete structures, 

 is by direct large-scale and model experiments and by that type of co- 

 operative action of the scientist and practical engineer which has achieved 



L 2 



