RAILS. in the decade from 1880 to 1890 but few changes have ve taken place in the theory or practice of construction and maintenance of the permanent way of railroads. One im portant change has taken place, in the United States at least, in theory, and to sonic degree in practice. That relates to the form, weight. and composition of rails. The iron rail no longer exists except as a relic. In 1880 there were in the tracks of the United States, 70.741 miles of iron rail, and 37,329 miles of steel (Tenth Census). At the end of 1890 there were still 40,700 miles of track laid with iron and 167,600 miles with steel. (Poor s Manual.) The question of steel or iron rails was settled long before 18811, and, in fact, commercial roil ing of steel rails began in the United States in 1867. The important change of the last decade has been in the steel rail itself. In 1880 rail makers and railroad engineers had begun working on the theory that a comparatively soft rail would wear better than one con siderably harder. Accordingly, rails were made with about 0 30 per cent. of carbon. The influence of this theory became still more marked by 1885 or 1886, and indeed the doctrine has not yet been absolutely disproved, but it has been shown to be so improbable. that the hardness of rails is being increased quite generally. The practice in the United States now is to use 0.10 to 0.00 per cent. of carbon, according to the weight of the section, with a ten. dency to 0.50 or 0.55 as an averlige. The most recent example of a heavy rail, designed to be high in carbon and stiff in section, is the Boston & Albany Railroad Co.'s rail, 95 lbs. per yard, rolled by the Bethlehem Iron Co., in 1891. This is the heaviest rail used in the United States up to t he end of 1891. (A rail weighing 100 lbs. per yard has been laid in the St. Clair tunnel, Grand Trunk Railway.) This Boston & Albany mil is important as an example of late and good practice in composition and design. Its general outline is shown in Fig. 1. The chemical specifications call for carbon 0.60. and phosphorus not to exceed 0.06. per cent. Physical tests give an elastic limit of 55,000 to 60,000 lbs. per sq. in., and from 12 to 18 per Cent. elongation. In England the percentage of carbon has long been about 0.40, and in France it is much higher. Rails above 0.60 per cent. carbon are common there, and they often run as high as 1 per cent. The theory of the better wear of very soft rails never affected steel rail practice so much in France as in the United States.
The change in the the my of the section is shown by Figs. 2 and 3. These are 85-lb. rails rolled for comparative trial. Fig. 2 shows, in a general way, the best section according to the theories of 1880 ; Fig. .3 shows the theories of 1890. It must he borne in mind that the later form is still tentative. The earlier section was adopted to get the mass of metal in the head of the rail to take the wheel wear, while the web and flange were re duced to the minimum dimensions which would give reasonable bearing on the ties. endurance against corrosion, and vertical stiffness. The result was disappointing. It gradually appeared that the rails with large heads did not weal' as long as rails of earlier make, with smaller heads. even when these last were of considerably lighter section. The investigations of engineers, rail makers, and students have gradually crystallized into the doctrine that the mass of metal in a steel rail should be disposed not merely with regard to wheel wear and to get stiffness as a beam, but so that the metal in the head ;:hall he thoroughly worked by the rolls, and that the cooling shall be uniform through out the section. as nearly as may be. In the type of section shown by Fig. 2, the distribu tion of metal is about : Head, 47.51 per cent.; web, 18.95 per cent.; flange, 33.54 per cent. In Fig 3 the proportions are : Bead, 41 per cent.: web, 21.46 per cent.; flange. 37.51 per cent. In the latter section the metal in the head, although less in mass, is better compacted, and defects in the ingot are inure likely to be worked out : besides, cooling strains are less, and less straightening of the rail is necessary in the mill. The more modern theory appears to be borne out by facts. but some veers must pass yet before it is absolutely demonstrated to be correct. (See for discussions of these matters, Trans. A. S. C. E., 1888 to 1801 ; Trans. Ain. Inst. of Mining Engineers, .1888 to 18111 ; and technical journals, especially the Rail road Gazelle, 1886 to 1891.) The average weight of rails rolled in the United States in 1891 is estimated by the makers at between 65 and 70 lbs. per yard, but many have been rolled of 75, 80, and 85 lbs., and some of 90 and 95 lbs. There is no means of making an accurate estimate of the average weight rolled in 1880, lint 67 lbs. per yard may be taken as the maximum of that (late, while 56 lbs. was a very common weight.