These show that the average iron contents of 100,000 tons of New Bed ore (65.12) coincides closely with the second low iron contents of Lover's Hole ore (65.10), but the average from two other workings only, viz., Big Pit and Little Pit, exceed this second low analysis. None of the averages, however, approach the lowest analysis of Lover's Hole ore (61.20). The richest ore from the various openings, as shown from the analysis, is below the average of Lover's Hole opening (68.60) in Orchard Pit and Roe Shaft only. The phosphorus average of Lover's Hole ore (0.033) exceeds that of four of the openings and is greater than the average phosphorus of all the openings. Comparing these as to the amounts of phosphorus in one hundred parts of iron we have: average of Lover's Hole ore, .0481 per cent.; average of all New Bed pure ore, .0444 per cent.

* Phosphorus contents in this column are from the same samples in which the iron is determined.

A SYSTEM OF RAIL-SECTIONS IN SERIES.

BY P. H. DUDLEY, NEW YORK CITY.

(Washington Meeting, February, 1890.)

A QUARTER of a century of service of steel rails on our oldest railroads, many of which have changed their standard sections three or four times, has furnished, and is furnishing, excellent opportunities to study their wear and behavior in the track. Such experience is valuable for the design of new sections, now required to meet the constantly increasing severity of railroad service.

Before presenting the new sections for consideration, I will briefly call attention to some features of past experience.

For convenience of study, rails may be separated according to their weight and height into four groups, viz.:

1. Rails weighing per yard from 52 to 60 pounds, and 34 to 4 inches high.

Experience in this country is confined mainly to sections in groups Nos. 1, 2 and 3.

During the past few years I have had the opportunity to run my car over many thousands of miles of railway laid with the principal sections in use, and to obtain diagrams showing their deflections, permanent set, and, in a great measure, their manner of wear.

The First Group.-The first steel rails imported, and those manufactured here, were light sections-included in the first group-and many of them, being rolled in rolls used for rolling iron sections, have pear-shaped heads, 23 to 21⁄2 inches wide, the base being heavy and from to inch narrower than the height.

These rails were laid upon ties of solid timber, at least 8 inches face, spaced 2 feet from center to center. The moments of inertia of these rails were low, only ranging from 8 for the lightest sections to 12 for the heaviest. The deflections were, of course, great; but the quality of the metal was excellent. It was physically hard and

tough, did not easily take a permanent set and wore very smooth. The loss of metal per million tons of traffic was, for Sir John Brown & Co.'s rails, on tangents, .01 to .015 pound per yard; on gradient tangents, .015 to .02 pound per yard; on gradient curves, .025 to .03 pound per yard. The rails of Charles Cammell & Co., Landore & Siemens and the Barrow Company were quite as good. Several of the American brands were excellent, and, by many railroad officials, considered equal to the imported.

The wheel-tonnage, however, was much lighter than at present, being 4 to 6 tons per locomotive-driver; 2 to 2 tons per freight

FIG. 1.

Ashbel Welch
No. 28

About 572 lbs.per yard.

car wheel; 3 to 33 tons per passenger-coach wheel. The cutting of ties under these rails was rapid even for the light tonnage.

The Second Group.-Observing these features led to the design of stiffer and heavier sections, in Group No. 2 of the Ashbel Welch type, of 60, 63, 65 and 67 pounds per yard. His design was made

in 1866, and by 1870 several sections were in service. These rails were from 4 to 4 inches high, had heavy webs and bases, the latter usually to inch less in width than the height of the section. This type is shown by Fig. 1.

The metal in many of these sections was intended to be fully equal to that of the earliest rails; and it wore well, though, as we

FIG. 2.

67 lbs.

Steel

now know, the loss of metal should be greater per million tons of traffic, than on the lighter and more flexible rails.

Many thousands of tons of these rails were laid by the trunklines, some of them being still in service, while a large part was removed, to be replaced with heavier sections, on account of increased traffic, and not because the rails were entirely worn out.

The satisfactory observed wear of these rails led to the conclusion that the only wear was upon the head, and, therefore, that the governing principles of new designs of rail-sections should be a shallow base as wide as the height, and a thin web supporting a narrow and deep head for wear. After 1873, these principles were adopted in practice, and millions of tons of rails of this type, represented

FIG. 3.

In 1881, Out 1889.

by Fig. 2, were put into the track (most of the sections being included in Group No. 2) prior to the year 1880.

In many cases, where the change was made from one type of section to the other, the weight per yard remained the same. These being put in the same tracks, carrying yearly the same tonnage, have furnished excellent opportunities to study the comparative wear of