6o8 



NA TURE 



[OcTOIiER 2 2, 1896 



\M.uM ;ict, the efi'fcl of continued work done on the surface is 

 the sjiLiwth of compressive stress exceeding; elastic resistance. 



Ill the case of railway rails the freedom for the flow of the 

 nialerial is very limited. Hardening of the surface takes place, 

 ami destructive compres.sion of the surface material is set up. 

 If the material be cast iron, the destructive compression causes 

 crumliling of the superficial parts, and the consequent relief of the 

 material immediately below it from stress beyond that of elastic 

 compression ; but when the material is that of steel rails, the 

 stress accumulates, the upper part near the surface being under 

 intense compression, differentiating from a maximum at the 

 surface. This compression gives rise to molecular stresses, 

 analogous to those which, on the compression side or inner curve 

 (if a bar bent on itself, originate traverse flaws on that side. 

 This condition of compression exists along the whole length of a 

 rail, so that when its magnitude is sufficient to originate crumli- 

 ling ur minute flaws, any unusual impact stress, or a stress in the 

 direction oppo.site to that brought about by the usual rolling load, 

 the rail may break into two or into numerous pieces. Stresses 

 originating in the same manner explain the fracture of railway 

 tyres as described fully by the author in the " Proceedings of the 

 Institution of Civil Engineers,'' 1876, vol. xlvii. 



A good discussion followed the reading of this paper. It was 

 opened by Prof. Unwin, who took a somewhat different view 

 from that of the author. The latter, the speaker pointed out, 

 attributed the ultimate failure of a rail to the number of trains 

 which passed over it : but his, the speaker's, experience told him 

 that there was most danger in new rails. Again, according to 

 the paper, one would expect soft rails to give way more quickly 

 than hard ones ; but here again experience negatived the assump- 

 tion. A defect in the paper, however, was that the author had 

 neglected to consider the composition of the rail, and this was 

 the governing factor. Other points to which the speaker made 

 reference were fatigue, and the change from a homogeneous 

 to a non-homogeneous material. Ii was well known that a rail 

 might be used in one way for a considerable time, but that when 

 turned over it would be liable to break, and the speaker 

 further illustrated his point by the analogy of a punched hole ; 

 but in this case one part was put in tension, so that annealing 

 removed the defect. These things, however, did not solve the 

 problem, and in his opinion work put upon the rail in use 

 .strengthened rather than weakened it ; but the initial condition 

 had far more influence than the rolling of wheels. 



Mr. Johnson, of the Midland Railway, had strips taken from 

 various parts of broken rails, and did not find difference in com- 

 position. Fractures had undoubtedly occurred through rails 

 being made from a " piped" ingot — that is to say, one in which 

 the whole of the head and pipe, in which the impurities collect, 

 had not been sufficiently removed. 



Dr. Anderson, Director-General of Ordnance Factories, 

 pointed out the similarity of the effects described by the author 

 in the case of rails and those observed in big guns which had 

 been much fired. In the bore of guns a large number of minute 

 cracks were discovered, and the deterioration of the A tube of a 

 ^un was due to the powder gases breaking out the squares. 

 Here there was ultimate compression and release of pressure, as 

 in a rail. 



The President had Examined rails which had failed, by the 

 microscope, and had noticed the minute cracks referred to. He 

 would point out that rails often gave way at the ends, and this 

 bore out the theory that defects were caused by insufficient 

 cro|iping leaving the "pipe." He pointed out that a crack 

 once started might easily be extended by lower strains than 

 would be required to start it ; just as a tear or rent commenced 

 on a piece of paper would be easily continued. Prof. Hele- 

 Shaw pointed out that if the rail were planed, the latter 

 defect would be removed. It may be worth putting on 

 record that the late Mr. Spooner, chief engineer to the 

 Kestiniog Moimtain Railway, who used to turn his rails 

 at times, once told us that an unplaned rail was more liable 

 to break than one which had fjeen planed. Of course the 

 object of planing was not undertaken with a view to prevent 

 breakage, but to take out the dents from the chairs ; Ijut the 

 result stated had been observed. 



In replying to the discussion Mr. Beaumont stated, in regard 

 to Prof. Unwin's remarks, that he, the speaker, had submitted 

 facts, and not speculations, to explain breakage of rails. 

 The Board of Trade inquiry on the subject had proved that there 

 was a good deal to learn, and he had mainly put forward his 

 paper with a view to raising di.scussion. He could produce 



figures tending to show that sometimes the hardest rails lasted 

 longest, though when they did give way thej' were apt to break 

 into a greater number of pieces. Undoubtcdl)' the rail must be 

 of good steel — not impure — to do its work properly: that, he 

 had concluded, was a foregone conclusion. The question of the 

 rail forming a continuous girder aflected the matter of end 

 break.age, and in this respect the influence of the modern stift 

 fish-jjlate had to be considered. 



On the following day, Friday, .September 18, the proceedings 

 opened with the report of the Sectional Committee appointed to 

 consider the effect of wind and atmospheric pressure on the 

 tides. The members of the Committee were Profs. L. F. 

 Vernon-Harcourt and W. C. Unwin, Messrs. d. F. Deacon and 

 W. H Wheeler. The latter acted as secretary, and drew up 

 the report. Information had been obtained from various ports in 

 England. It was concluded, firstly, that the tides are influenced 

 both by atmospheric pre.s.sure and by the wind to an extent 

 which considerably affects their height ; .secondly, that the 

 height of about one-fourth the tides is affected by wind ; thirdly, 

 that the atmospheric pressure affecting the tides operates over 

 so wide an area, that the local indications given by the baro- 

 meter at any particular spot do not aflord any trustworthy guide 

 as to the effect on the tide of that particular port ; fourthly, that 

 although, so far as the a\'erage results go, there can be traced a 

 direct connection between the force and direction of the wind 

 and the variation in the height of the tides, yet there is so much 

 discrepancy in the average results when applied to individual 

 tides that no satisfactory formula can be established for indicating 

 the amount of variation in the height of the tide due to any 

 given force of wind ; fifthly, the results given in the tables 

 attached to the report relating to atmospheric pressure indicate 

 that the effect of this is greater than h.as generally been allowed, 

 a variation of \ inch from the average pressure causing a varia- 

 tion of 15 inches in the height of the tides .\s the report will 

 be printed in full in the published Pi-oaediiigs oi the Associa- 

 tion, we have thought it unnecessary to give more than the 

 conclusions reached, but the whole is well worthy of the 

 attention of those interested in the subject. Mr. Wheeler is 

 well known as a trustworthy and diligent student of this 

 question, and his professional status enables him to obtain 

 information from a wide source. 



A brief report on the calibration of instruments in en- 

 gineering laboratories was the next item in the programme. 

 Copies of this report, so far as we could ascertain, were not 

 distributed. 



Mr. Barry's lecture on the Tower Bridge followed, and 

 attracted a large audience. It was interesting, and the lantern 

 slides were well managed. Mr. J. Parry followed with a long 

 paper of the historical-record order, dealing with the Liverpool 

 Watersvorks. The last item on this day was a contribution by 

 Mr. A. J. Maginnis, entitled "The present position of the 

 British North Atlantic Mail Service.'' It was a good paper in 

 its way, but its way was not quite that of mechanical science ; 

 indeed, the author dwelt rather on the economics of ocean 

 service than on its engineering aspects. Some instructive 

 figures in regard to coal consumption were given, it being stated, 

 among other things, that the Campania burns 20 tons of coal 

 per hour. To drive an improved Canifiania, 700 feet long and 

 74 feet wide, 23 to 24 knots would require 46,000 indicated 

 horse-power, supposing existing practice were followed. The 

 cost of the vessel would be ^800,000. 



The next sitting of Section G was held on the Monday 

 following, September 21, and was, according to custom, devoted 

 to electrical engineering. The first business was the reading of 

 a report by the Committee on small screw gauges. This report has 

 been looked forward to with interest for some time. It will be 

 printed in full in the Proceedings of the .Association. Mr. 

 Preece (the chairman of the Committee) drew up the report. 

 After giving details of the method of work followed by the 

 Committee, and referring to the labours of others in the same 

 field, the report proceeded to notice a method, suggested by 

 Colonel Watkin, for making very accurate comparison. There 

 would be thrown, side by side on a screen, photographic 

 images of the screw to be examined, and of the standard with 

 which would be compared, together with the image of a scale 

 which might be divided to one ten-thousandth of an inch. 

 The images of these three objects being so close to one 

 another, a comparison to a very high degree of accuracy 

 could be made. Mr. Price, a member of the Committee, sub- 

 milted a microscopical method, in which the screw to be 



NO. 1408, VOL. 54] 



