TRANSACTIONS OF SECTION G. 985 



the ship started from Bristol in 1838, and arrived at New York in fourteen days 

 with 200 tons of coal in her bunlcers. 



Let me remind you of another somewhat similar instance of the way in which 

 the anxieties of enfrineers have been unnecessarily increased and jniblic alarm 

 gratuitously, though honestly, aroused. When the designs of the Forth Bridge — of 

 which the nation, and indeed the world, is proud — had been adopted both by the 

 railway companies who were to find the capital and by I'arliament, a most dis- 

 tinguished man of science^the then Astronomerlloyal — came to the conclusion that 

 the engineers had neglected certain laws which lie enunciated respecting the resist- 

 ing power of long struts to buckling, and that the bridge ought not to be constructed, 

 as he considered that, to use his own words, ' we may reasonably expect the 

 destruction of the Forth Bridge in a lighter gale than that which destroyed the Tay 

 Bridge.' All this was stated, no doubt, from a strong view of pidjlic duty, in a 

 letter to a public newspaper, though subsequently and frankly withdrawn. If 

 the bases of his calculations were right, the conclusion might have been correct; 

 but the fact was that there was no foundation worthy of the name for the 

 reasoning. Again, anotlier distinguished mathematician publicly criticised the 

 Forth Bridge with equal vigour, basing his views that it Avas fundamentally 

 incorrect on another set of equally erroneous assumptions, maintaining again that 

 it should not be permitted, because he proved by reasoning on those assumptions 

 that it must be absolutely unsale. 



Once more, in shipbuilding : until Mr. William Froude, some years prior to 1875, 

 made his experiments by means of models on the highly difficult and otherwise 

 almost insoluble causes of the retardation of ships and their behaviour in waves, 

 beginning at the beginning, taking nothing for granted, and eliminating all ele- 

 ments of possible errors, little or nothing was known of the laws governing these 

 questions. Laws had been laid down by high authorities as to the causes of retarda- 

 tion of ships, many of which, in fact, were not true, while some of the assigned 

 causes were non-existent and some real causes were unrecognised. Mr. Froude 

 was told that little or no information could be learnt from experiments on models 

 which would be applicable to full-sized ships, and that ships must continue to be 

 designed and engines built on data which, scientifically speaking, were assump- 

 tions. The outcome has been that Mr. Froude 's a prion depreciated experi- 

 ments with models have solved most of the questions relating t<> that branch of 

 naval architecture; and at the present time every ship in the Royal Navy, 

 and not a few in the merchant service, are designed in accordance with the data 

 so gained. 



Another example of hasty generalisation occurs to me, and that is on the 

 important question of wind pressure. Tredgold, who undoubtedly was one of th& 

 soundest of engineers, laid down in 1840 that a pressure of 40 lbs. per square 

 foot should be provided for ; reasoning, no doubt, from the fact that such a pressure 

 had in this country been registered on a wind gauge of a square foot or less in area. 

 As a consequence, he assumed that the same force could be exerted by the wind on. 

 areas of any dimensions. Thus, roofs and bridges, wherever any calculations of wind 

 pressure were, in fact, made, were designed for a pressure of 40 lbs. per square foot 

 of the whole exposed surface ; and under the alarm caused by the fall of the Tay 

 Bridge in 1879, the piers of which were not, probably, strong enough to resist a 

 horizontal pressure of one-fifth of such an amount, a further general assumption 

 was made, and railway bridges throughout the kingdom were ordered by the Board 

 of Trade in 1880, acting no doubt on expert advice, to be in futuie designed, and 

 are designed to this day, to resist 56 lbs. of horizontal wind pressure on the whole 

 exposed area with the ordinary factors of safety for the materials employed, as if 

 such horizontal strain were a working load. 



It had, for a long time previously to this order of Government being issued, been 

 suspected that these small-gauge experiments were untrustworthy, and subsequent 

 experiments at the Forth Bridge on two wind gauges of 300 square feet and of 

 1^ square feet respectively indicated that with an increase of area the unit of 

 pressure fell off in a very marked degree. Under the same conditions of wind 

 and exposure the larger gauge registered a pressure 38"7 per cent, less per square 



