144 TRANSACTIONS OF THE [FEB. 1, 



see where the deflections occurred, which was not only a great 

 educator to them, but directed the work of the latter to tbe 

 places needing most attention. 



The diagrams for 1882 repeated all the striking characteristics 

 of those for 1881, not an important curve per mile on the con- 

 densed diagrams having been reversed unless new rails had been 

 laid. The permanent set in the old rails had not been reduced. 

 The joints were down and weak. The receiving ends of the rails 

 were cutting out. The rails were loose on the ties moving up 

 and down for every passing wheel, cutting the ties, disturbing 

 the ballast and roadbed. The dynamic effects of the wheel loads 

 were so great that the tracks could not acquire a high degree of 

 stabilit^^ Thoughtful railwa\' officials accepted the fact that the 

 rails were not stiff enough for the traffic, and, after studying the 

 diagrams of several thousand miles of track, I designed the five- 

 inch 80-pound steel rail for the New York Central and Hudson 

 River Railroad in April, 188.3. With 23 per cent, more metal 

 than in their four and one-half inch 6.5 pound rail, I increased 

 the stiffness 6fi per cent. The rail was rolled and put into ser- 

 vice in 1884. This was the first five-inch steel rail laid in the 

 United States. 



Having noted the wear of the steel on all t3'pes of the heads of 

 rail in the tracks, I made it broad and much thinner than usual, 

 which has since become the type for modern sections. 



The five-inch 80-pound rail once in the track demonstrated 

 the great value of stiffness in maintenance of way and vvas fol- 

 lowed by a number of sections of 80 and 85 pounds per yard in 

 the leading railroads oP'the country. 



This led at once to the introduction of larger and heavier 

 locomotives and cars, and faster and heavier trains. 



The undulations in the track reduced to about four feet per 

 mile on the first five-inch 80-pound rails. The five-inch 80-pound 

 rails were all straightened at the mills on narrow spaced supports, 

 used for much lighter rails, and were indented by the gag wher- 

 ever applied, giving the surface a series of minute waves, which 

 produces from one to one-half feet of the total undulation found 

 upon these rails per mile. 



These undulations increase the d3'namic effects of the wheel 

 loads, cutting out the ties, disturbing the ballast, besides giving 

 a very unpleasant tremor to the cars when passing over them. 



In 1883, after I had designed the five-inch 80-pound rail, but 

 before any were rolled and in the track, I calculated what I 

 considered would be the possible condition of track for such 

 rails when finished smooth, and put it on the diagram. The re- 

 sults seemed so impossible that I was probably the only one for 

 many years who expected to see them realized. 



