three assistants, one engineer and two assistants remaining for the general service. The staff was subsequently increased by one party of similar constitution, and by six assistants, giving two to each party. Owing to sickness and various causes there were generally only three or four of these assistants available.

For the preparation of a plan and longitudinal section, the work was divided into: a. General reconnaissance. b. Rough survey. c. Detail survey.

The general reconnaissance was performed by means of aneroid and compass. The former instrument was also used for the altitudes of the general survey, whilst the position of the stations was determined by "tranche-montagne" compass and chain. Each of the three parties was provided with two Richer's tacheometers. As the work progressed, and the surveyors became more practised, two staffholders per instrument were found to be insufficient, and this number had to be increased to six. More are needed in a broken or wooded country than in open plains.

The mean level of the sea was determined by three tide gauges in Troesan Bay, and the result carried over to Solok by a Troughton and Simms level, and from thence on through the Highlands by the tacheometer. A flying survey with the aneroid, chiefly without datum, over an extensive region gave an outline of the country, and pointed out those valleys likely to come into consideration in the laying out of railways. A barometric levellingfollowing closely the line of main watershed indicated the passes. by which the country remaining to be surveyed was limited. The tranche-montagne" compass proved a useful instrument, being very handy, rapidly set up, and giving fairly long sights of 80 to 100 yards. Distances were measured by the chain, and verified by steel tapes.

The general survey was plotted to a scale of 1 to 10,000, and, with the aneroid readings added, was forwarded to the head office and there worked up into a complete whole. On these plans a direction was drawn by the chief of the party, to be afterwards modified by consideration on the spot by the head of the survey. These maps formed the foundation of later detail surveys, and show not only the mountain passes and general direction of the railway, but also the maximum and minimum gradient permitted by the ground.

For a complete survey for a railway from a mountain pass to a valley or plateau, a triangular piece of country must be taken, of which the mountain range is the base and the two sides are the centre lines of two railways, one of which has the maximum and the other the minimum ruling gradient. Within this triangle the best direction for the line must be found. The cost of such a survey confined the Author to the survey of one broad strip of country.

On the maps of the preliminary or general survey no contours are sketched, but for the elucidation of the sketch maps single contours were often drawn. The detail survey was executed

exclusively with the tacheometer, and was plotted to a scale of 1 to 2,000. Principal stations were transferred to the plan by means of co-ordinates, for which purpose the Author recommends ruled "sectional" paper. Whenever two or more tacheometer parties could work in the same direction, the base line was set out and measured in advance by the chief of the party, and the surveyors started at distances apart of 2,000 to 3,000 yards.

The portion between the secondary watershed and Brandewijn Bay presented extraordinary difficulties; it was here impossible to work through any length of base line at once, and as with the utmost zeal it was found that the plans could not be sent in by the required date, two parties were set to work so as to meet. The result was satisfactory, notwithstanding that for two days before meeting the parties could only communicate by a long detour.

In order to afford a check upon the progress of the surveys the bi-monthly reports to head-quarters contained a statement of the latest points observed by the various parties. These points were laid down on a general map provided with contours corresponding to those used by the surveyors.

The survey for the Padanj Highland railway was carried out in a similar manner. In this beautiful and well-cultivated region the open character of the country enabled the survey to proceed very quickly. It was necessary, in consequence of the breadth of the valleys, to extend the width of country included in the flying survey to 3,300 yards, and in some cases even 5,500 yards. Here also the "tranche-montagne" compass was used, together with aneroid and chain. The Author, however, advises the use of the tacheometer even in flying survey, whenever, as in the case of a watershed, a considerable number of points can be observed at one setting up. It is particularly useful in a country under cultivation, where damage would arise from the trailing of the chain. On this flying survey, plotted to a scale of 1 to 10,000, was based the tacheometer detail survey, which, as in the case of the coal line, was conducted throughout by only two persons per

instrument.

The main roads and some bye-roads were surveyed with the tacheometer; this survey was afterwards extended, so that a network of triangles was obtained whereof the sides have been surveyed with the tacheometer, and which has formed the basis of the map of Central Sumatra accompanying the second report (scale 1 to 100,000).

One assistant in fifty-two days (including interruptions from rain) levelled and brought on to plan 62 miles of road, with an average of 9 stations and 25 neighbouring points observed per mile, in which he was assisted by one native ganger and eight coolies. The accuracy was satisfactory, and greater than could have been obtained by the use of any other instrument. The method used at Sumatra is comparatively cheap. There have been surveyed in full detail, planned and estimated-for the colliery line, 101 miles; in the Highlands, 82 miles; total, 183 miles.

COAL DEPOT ON THE PHILADELPHIA AND READING RAILWAY. 309

Above 218 miles of main roads have been tacheometered, plotted, &c. There were employed under the Author in one month of railway surveying one hundred and seventy-five engineers, two hundred and sixty Dutch assistants, sixty native assistants, equal to the labour of three complete parties during twenty-three months. Taking five months' labour of one party for the survey of harbours and such incidental work as sounding and boring, &c., sixty-four months of one party have been spent on the railway

surveys.

Dividing the time between the colliery railway and the Highland railway in the ratio of 4 to 1, it appears that per month and per party there was surveyed, plotted, and estimated, of the colliery railway 2 miles, of the Highland railway 6.43 miles.

The total cost per mile is for the colliery line £404; for the Highland roads, £125, which compares favourably with the costs of similar works executed in Europe.

A. S. M.

Coal Depot on the Philadelphia and Reading Railway.
By H. BLANCK.

(Zeitschrift fur Bauwesen, vol. xxvii., pp. 427-440, 547-560.)

The Author in an article on the construction of American railways gives a variety of details about the permanent way. While on the subject of rails, he states that their usual length is 24 feet; but so many advantages have been found to result from the use of rails 30 feet long, that the latter are now almost universally adopted. At the works of the Philadelphia and Reading Company old rails are converted into new ones; the entire make in the year being 30,000 tons, and the proportion of old to new rails two-thirds. The Author gives a description of the process, adding that the results are so satisfactory that other railway companies are introducing the system. An account is given of the above Company's coal depôt at Richmond, near Philadelphia. This immense establishment is situated a short distance from the Delaware, and is used almost entirely for the shipment of coal. The Company possess two rich coal-fields in the Schuylkill district, separated from each other by the Broad Mountain ridge, and connected by an unfavourably situated tunnel, with a single way. As this was far from sufficient for the traffic, communications have been established with the Mahoney fields, across the ridge. The height intervening is 360 feet, and the length of the inclined plane, which forms a parabolic curve, is 2,350 feet. The inclination is, therefore, about 1 in 6. There are two pairs of rails of ordinary gauge, 14 feet apart from centre to centre, the ascending and descending trucks being on the road simultaneously. The rails rest on longi

tudinal sleepers, and these again on cross sleepers. On the latter there is an intermediate pair of rails, 3 feet 2 inches apart, and between these, at intervals of 50 feet, guiding pulleys about 1 foot in diameter. The trucks are pushed up, or held back, by small four-wheeled "barneys," to which the steel rope is attached. In order not to interrupt the traffic, the "barneys," when they arrive at the lower end of the inclined plane, run into a depression between the outer rails; the rope, by means of a shifting horizontal wheel, being here attached to a counterweight of 18 cwt., calculated to keep an equal strain between the two ropes while working. The rope is worked at the top by a 500-HP. engine. Ten loaded and fifteen empty trucks can be worked at the same time, at the rate of 20 miles an hour, making a total of two thousand two hundred trucks, at 5 tons of coal each per day. When 2,000,000 tons have been worked, the ropes (1 inch and 2 inches in diameter) are replaced by new ones, even if not worn out. At the lower end of the inclined plane the lines branch out into a fan-shaped network of nearly fifty pairs of rails. These are laid on an embankment about 13 feet high, to a point not far distant from the docks; from this point they are fixed on framings of massive timber, which are planked below, so as to give storageroom for 180,000 tons of coal. There are twenty-one wharves, at which one hundred and twenty vessels of all kinds can load simultaneously. The annual shipment of coal has hitherto not exceeded 2,500,000 tons, but with the means at disposal now from 3,000,000 to 4,000,000 tons can be shipped.

J. J. W.

Swedish State Railways. By P. W. ALMGUIST.

(Ingeniörs-Föreningens Förhandlingar, Stockholm, 1877, pp. 7-20.)

In the Traffic-reports on the Swedish State railways the receipts are given under the two headings of passenger and goods traffic, while the expenditure appears under four, viz., traffic, permanent way, machinery (locomotive and carriage), and office expenses. Hence it is impossible to ascertain directly from the reports what proportion of the net revenue each kind of traffic contributes. The Author gave this information for 1865, and now compares the results obtained for that year with those for 1875.

The hypotheses on which he bases his calculations are—

1. That the cost of working and repairing each class of wagon or carriage is proportional to the train mileage run.

2. That locomotive, traffic, and permanent way expenses are chargeable against each class of traffic, proportionally to the train mileage run by the wagons of that class, and to the gross weight moved.

3. That office expenses are chargeable to each class in the proportion of the sums expended under the other headings.

On the same hypotheses he ascertains also the receipts and expenditure per train mile as under :

:

It also appears, that each engine drew on an average 2·37 times its own weight in 1875, as against 2.34 times in 1865; that the proportion of dead weight to paying load was, on an average, for passengers as 17 to 1, and for goods as 2.66 to 1; and that the paying load averaged only two-thirds of the weight of the engine and tender, or one-fifth of that of the whole train by which it was carried.

M. L.

On the Duration of Steel and Iron Rails.

(Zeitschrift des berg- und hüttenmännischen Vereines für Kärnten, vol. ix., pp. 341–345.) The following figures refer to the main line of the CologneMinden railway, which has a total length of way of 1,357 miles, or double that length of rails in use, exclusive of colliery