use of steam locomotives. Indeed, in 1830 a small engine was built by Peter Cooper and made experimental trips on this road, but the first locomotive to be put in actual operation was installed in 1831. In 1831 the De Witt Clinton, a locomotive built by the West Point Foundry, was put into service on the Hudson and Mohawk Railroad. The next railway to mark a step in the development of the railway system of the United States was the Camden and Amboy Railroad, begun in 1831 and completed from Bordentown to South Amboy, N. J., 34 miles, in 1832. The president of this road, Col. Robert L. Stevens, conceived the idea that an all-iron rail would be preferable to the iron-strapped wooden rails employed on all previous American roads. There was no rolling mill in America capable of rolling such rails, however, and Mr. Stevens went to England to secure them. His request of the English ironmasters was for a rail having a head similar to that then in use upon the principal British roads, but with a wide flat base to the web, which he proposed to secure to the supporting blocks or sills by hook-headed spikes. Considerable difficulty was experienced in getting this request fulfilled, but in May, 1831, the first 500 rails, 15 feet long and weighing 30 pounds per yard, reached Philadelphia, and were placed in the track, thus recording the first use of the flanged T-rail, which has since become universal in America and is extensively employed abroad. It is important to note here that the flanged T-rail was reinvented in England in 1836 by Mr. Charles B. Vignoles, and that rails of this form are known abroad as Vignoles rails. Mr. Stevens also invented the fish plates and the hook-headed spike.
Table III.— Showing the Number of Miles of Railway Constructed and in Operation by Decades, in the United States, from 1830 to 1900 Inclusive
| YEAR | Miles in Operation |
| 1830 | 23 |
| 1840 | 2,818 |
| 1850 | 9,021 |
| 1860 | 30,626 |
| 1870 | 52,922 |
| 1880 | 93,262 |
| 1890 | 166,654 |
| 1900 | 194,321 |
Table IV.— Showing Mileage of Various Classes of Railway in the United States on June 30, 1900
| CLASS OF TRACK | Miles |
| Single track | 192,556 |
| Second track | 12,151 |
| Third track | 1,094 |
| Fourth track | 829 |
| Yard track and sidings | 52,153 |
| Total track | 258,784 |
Table V.— Showing Number of Each Class of Railway Cars in Operation in the United States on June 30, 1900
| CLASS OF SERVICE | Number |
| Passenger | 34,713 |
| Freight | 1,365,531 |
| Company's | 50,594 |
| Total | 1,450,838 |
Railway Surveys. The surveying operations requisite to and preceding the construction of a railway are in general a reconnaissance, a preliminary survey, and a locating survey or location. The reconnaissance is a general and somewhat hasty examination of the country through which the proposed road is to pass for the purpose of noting its more prominent features and acquiring a general knowledge of its topography with reference to the selection of a suitable route. A preliminary survey consists of an instrumental examination of the country along the several available routes for the purpose of obtaining such details of distances, elevations, and topography as may be necessary to prepare a map and profile of each, make an approximate estimate of the cost of the road, and furnish the data from which definitely to locate the line. The locating survey consists specifically in establishing the centre line of the road on the ground in the position which it is finally to occupy. Defined more broadly, the location consists first in choosing the best route from the several which are available, and second in selecting for the chosen route the best combination of grades and curves. In determining the best combination of grades and curves for the route chosen the engineer has to take into account both the cost of construction and the cost of operation. On one hand he has the annual interest upon the original cost, and on the other the annual expense of operating the road. But the best combination of grades and curves is that which will render the sum of these two a minimum. To select the best line from several available lines, the engineer determines the most economical combinations of grades and curves for each one, calculates the interest on the entire cost of constructing the line with this combination and also the annual expense of operating the line, and takes the sum of the two amounts. That route is best in respect to which this sum is the least. In this last statement it is assumed that so far as their ability to command traffic is concerned all of the available routes are on a parity. This is not always the case. Sometimes one route is superior to any of the others in its ability to command traffic while being inferior in its ability to present the most economical combination of grades and curves. In such a case it often becomes the duty of the engineer to select the more expensive route for the sake of securing the greater amount of available traffic. It is plain upon very little thought that to answer each of these broad general questions a multitude of minor factors have to be carefully integrated, and that altogether the location of a railway is a task which, if it be well performed, calls for skill, experience, and good judgment on the part of the engineer. Beyond the statement of this fact it is impossible to proceed within the limits of the space available in this article, but the reader who wishes to study the problem of railway location in detail will find it presented at great length in Wellington, Economic Theory of Railway Location (New York, 1900).
When the engineer has chosen his route and has selected a combination of grades and curves for this route his next task is to establish its centre line on the ground with all the grades and curves properly indicated. In plan the centre line consists of a combination of straight lines or tangents and of curves. The curves may be simple curves, that is, plain circular curves; or compound, that is, consisting of two or more circular arcs of different radii; or reverse curves, that is, two simple curves so joined as to form a curve like a flat letter S. Curves are further designated by their degrees of curvature. The degree of a curve is determined by the angle at
