leakage. In later years, however, a vertical hair-crack has appeared in the outer face, from top to bottom, plainly marked by a damp margin on each side. It can scarcely be said to be a leakage; but there is a slight oozing of moisture at the lower part of the wall. There is no sign whatever of vertical settlement, or of horizontal movement of the wall on either side of the fissure, and the Author is at a loss to account for its occurrence. It is presumed that the crack extends throughout the whole thickness of the wall, but this cannot be ascertained with certainty. The immense pressure of the water on the inside of the wall and the curvature of its plan no doubt keep the wall pressed against the abutments of this horizontal arch in perfect safety; and this leads the Author to approve of building dam-walls on a curved line, where they are bounded by hills only a few hundred feet apart. Mr. C. J. R. Williams, M. Inst. C.E., the Harbour Engineer at Christchurch, who has lately come from government service in Queensland, has told the Author that he is always prepared to find trouble in the use of concrete. It would be interesting to learn the opinions of other members of the Institution as to the choice of concrete or stonework in the construction of large dams, and also on the subject of straight and curved dams.

In completing the works of the Malmsbury reservoir, the Author was confronted with a new and somewhat difficult problem, the manner of solving which may here be described with advantage. The paving of the by-wash had been delayed, although the reservoir was kept filled with water nearly up to the level of the top of the dam, and a heavy flood caused the scour of a broad and deep gully, the extension of which for a few chains would have almost wrecked the reservoir. Without further delay the by-wash was paved with heavy blocks of masonry, and a concrete wall was thrown across the gully to prevent its further extension. This wall was curved in plan and its summit formed a vertical curve corresponding with the curve of the cross-section of the paved by-wash. The top of the wall, forming the bed of the coping, was horizontal in a direction radial to the curve of the plan, whilst the joints of the coping-stones were radial to the vertical curve of the top of the wall. After the building of the wall had been completed, it would have been easy to mark upon the top of the wall the position of each of the coping-stones, and to obtain by actual measurement the angles of the beds and joints, by which to cut templates to be sent to the quarry. But to obtain these templates simply from examination of the drawings was difficult and almost impossible. The following expedient was therefore adopted :-On an adjoining piece

of tolerably level ground the curve of half the length of the wall was set out, and at short distances apart pieces of scantling were erected and set to the intended batter of the wall. On these scantlings were marked the heights of the vertical curve of the top of the wall, to serve as a guide in nailing to them a light batten, which thus formed a working model of the top edge of the wall, from which the templates could be obtained almost as readily as from the completed wall itself. This plan proved perfectly successful, and the coping was afterwards placed upon the wall without the blocks requiring the slightest alteration. This problem illustrates the necessity for a practical acquaintance with the geometry of lines, surfaces and solids, as a part of the education of a Resident Engineer, who without this knowledge would be unable to give intelligent direction to the operations of a contractor, whose work it may be his duty to approve or condemn.

On the Author's return from Australia to New Zealand in 1876, when he entered into partnership with his son, Mr. Arthur Dudley Dobson, he found that great political changes had taken place since the time when he was appointed Provincial Engineer for Canterbury. The government of the country by provincial councils, each with a superintendent and executive, had come to an end. All legislative power was vested in the general Government, established at Wellington, with a House of Representatives, Legislative Council, Premier, and a Ministerial Executive. The provinces remained as geographical districts, with a certain amount of local administration of the Government departments, whilst public works were under the management of Road Boards, County Councils, and Municipalities, with, however, this great change, that the land fund, instead of coming directly to the Road Boards, went to the general Government at Wellington, and the funds for public works had to be raised by local rates assisted by Government grants. The effect of this to the Author's firm was that, whilst enjoying the freedom and independence of private practice, their sources of employment were many and the works entrusted to them were of a very varied nature. They made many extensive surveys, both public and private, and laid out new roads through pastoral country and many miles of water-races. These races required that the levels should be carefully contoured, in order to ensure uniformity of gradient and to avoid scouring-action, whilst the tunnels and sluices of the intake-works had to be skilfully designed and carefully executed. They also built many timber bridges, some of great length. The last of these was one of thirty-five 20-foot spans over the River Selwyn. The piles, capsills, and floor-beams were of iron-bark, imported from Australia,

whilst the flooring was of New Zealand heart of black birch. In addition, they carried out two important railway surveys for the Government.

This

The first of these was a length of 25 miles between Kaikoura and the River Waiau, on the line of the East Coast Railway now under construction between Waipara and Blenheim. was the roughest bit of survey-work that the Author has ever undertaken; some miles were along a rugged, broken coast-line, where the survey-party were often liable to be caught by the tide. Inland portions of the route were inaccessible to packhorses, and in a roadless country the distribution of 25 miles of heavy survey-pegs was a difficult matter. The Amuri district is broken up by deep, precipitous ravines, and the bush is of the usual dense character of the New Zealand forest. The provisioning of the camps was often difficult, and the severity of the weather made the work very trying, the survey-party often having their clothes frozen on their limbs after wading through the numerous streams. As in many places direct measurement would have been impracticable without ladders and scaffolding, the Author made use of tacheometry by means of the theodolite. Two different methods were adopted, namely, that of observing the intercept on the level-staff subtended by a given angle, and that of observing the angle subtended by a rod 4 feet 6 inches in length held up by one of the survey-hands. In these surveys frequent use was made of the aneroid barometer in filling up sectional details between two bench-marks, fixed either by ordinary levelling or by triangulation. Every set of observations was begun and closed by reference to a bench-mark; thus, if the height of the bench-mark at which work was commenced was 734 feet, the normal tabular reading of which was 29 17 whilst the actual reading was 28.99, the difference, 0.18, would be added as an index-error to the subsequent readings. If on arriving at the second bench-mark the index-error was greater or less, which often happened, as the barometer is seldom stationary, the difference was distributed to form a second revision of the barometer-readings, the time at which the readings were taken having been carefully recorded. The use of the aneroid barometer was found very helpful in running trial-sections. In order to run levels through the forest with the spirit-level it is necessary to cut lines to get a sight of the level-staff; but a barometer-reading can be obtained anywhere, although it may be necessary to crawl through the scrub on one's hands and knees.

The second of the two surveys referred to was over a length of 70 miles, between Waikari and Lake Brunner, forming one of the

routes proposed for the Midland Railway. This route was not ultimately approved of, that adopted being the one now under execution and partly completed, surveyed by the late Mr. C. Napier Bell, M. Inst. C.E., by way of the Otira Gorge, with a tunnel 6 miles in length under a summit-level of a little over 3,000 feet. The line surveyed by the Author lay along the valleys of the Hurunui and Teremakau Rivers, with a summit-level of considerably more than 3,000 feet, and, with the exception of a few miles of open country near Waikari, was entirely through river-bed and the usual dense New Zealand forest. But there were no engineering difficulties to be surmounted in the execution of the survey. The survey-party numbered twenty in all, with six horses, and there were no commissariat or climatic difficulties to contend with. The survey commenced at a trigonometrical station on the Canterbury Plains, where a true meridian was obtained from the reading of the registered bearing of another trigonometrical station in the secondary triangulation. The centre-line was staked out with branded pegs, 5 chains apart, on each side of which the topographical work was built up by ordinary survey methods. The work was plotted on separate sheets, imperial size, each sheet including a mile of survey, which was plotted to a scale of 3 chains to the inch, and on each sheet were ruled a meridian and an east-and-west line, on which were figured the distances west and north of the initial trigonometrical station. This enabled the work of each sheet to be transferred without the slightest difficulty to the Government maps. The work in the field was conducted by Mr. Arthur Dobson, and the mapping was done by the Author in the Christchurch office.

The Author is glad to learn from Sir Guilford Molesworth's Presidential address that interest is again being taken in railway gauges. He started railways in Canterbury in 1859 on the 5-foot 3-inch gauge, on which the lines were constructed for some distance beyond Christchurch. Then, on proceeding to Australia, he had the pleasure of managing the Melbourne suburban railways on the same gauge, with a heavy traffic in goods and passengers; and on his return to New Zealand in 1876 it was a bitter disappointment to find that the railway system of New Zealand was crippled by the adoption of the narrow gauge. He does not think that any of the advocates of a reduction of the standard gauge at all realize the evils of the narrow-gauge system. Apart from lightness of construction, the economy to be gained by reducing the distance between the rails is a very trifling factor in constructive cost. Diminution of trainspeed necessitates increase of the number of hands employed, and of the quantity of rolling-stock required; for example, an increase of

the train-speed from 15 to 30 miles an hour practically enables double the work to be done with the same staff and rolling-stock. And increase in the capacity of wagons means a great diminution in the length of the trains. Thus, if 200 tons of coal have to be taken in 10-ton trucks, twenty trucks will be required, whereas with 6-ton trucks the work cannot be done with less than thirty-three trucks, the 10-ton truck being very little longer than that carrying only 6 tons. On a level country, where the roads are all crossed on the level, the excessive length of the trains often leads to serious delay and inconvenience, especially at the large stations, crossed at each end by public roads. It is quite common for trains of forty trucks to run through the Lyttelton tunnel, and this number is often exceeded. The carrying-capacity of a single narrow-gauge line is very limited. The railways of Canterbury comprise more than 2,000 miles of line with a very heavy traffic in coal, firewood, timber, grain, sheep, wool, and frozen mutton, and serious difficulties are encountered for want of sufficient facilities for the crossing of trains moving in opposite directions. Increase in train-speed is promised; but, as the maximum accelerated rate will not exceed 25 miles an hour, the improvement will not amount to much, and it is impossible to reflect without bitterness upon the fact that, had the gauge been retained on which the construction of railways was first commenced in Canterbury, the express trains would be running at not less than 50 miles an hour, with all the advantages derivable from rapidity of transit.

A point of some importance in the New Zealand railway time-tables is that the height above the sea of the rail-level at every station is recorded. It is impossible to exaggerate the value of this extensive system of levels, all referred to one common datum, in surveys for drainage-operations or works of water-supply.