misleading, since this term had come to be applied to a vertical The Author. pipe used for a special purpose. The pipes referred to in the Paper were vertical and, of course, considerably above the ground; they were for the purpose of overcoming water-hammer, and not for producing pressure. The explanation given by Mr. Carruthers as to the reasons for the extra cost of the pumps over his original figures pointed more strongly, in the Author's opinion, to the doubtful expediency of installing much more expensive machinery; since if for two classes of plant the total working-costs (including interest, depreciation and working-expenses) were equal, it was better, and certainly less risky, to choose the plant entailing less capital expenditure. This would have been done had the Author been in supreme charge at the time when choice was being made. There was no doubt, however, as to the high duty of the plants installed. A complete Table showing all data, also the coefficients employed and the results obtained step by step, had been placed on record in the Institution, and it need therefore only be stated that the duty-tests had been most carefully made; moreover the Author had had the whole watched and concurred in by the Chief Mechanical Engineer of the State. Prior to the test-runs, all steam- and water-gauges, thermometers, weighing-machines, etc., were tested and their errors noted. The slip of the pumps was ascertained by three initial tests on different days, the water pumped each time being from a measured 10-foot zone of the suction-tank, and amounting to 310,220 gallons. The length of stroke was measured every 5 minutes, electric signals from the tank intimating to the engine-counter reader when the top and bottom respectively of the zone had been reached. Samples of coal from various parts of the heap to be used were taken, and their calorific values were obtained independently by means of Darling, Thompson and Carpenter calorimeters; and similar care was taken with the ash-pit residue. The state of the fires, and the height of the water in the gauges at the commencement of a test, were noted, and both were brought to the same condition by the end of the test. It had not been possible, owing to the way which the specifications were drawn, to follow in the tests exactly the lines laid down by the Institution, but had these been worked on, the duty obtained would have been still higher, for the air-pump discharges were run into a receiving-tank, and the temperature of the inflowing water was adopted as that of the condensed steam returning to the boilers in ordinary working course. There was no doubt that, as mentioned by Mr. Fitzmaurice, the servicereservoirs, specifically so termed, were small compared with what would be provided usually; but the Author had determined that

in

The Author. they should be sufficient at any rate to begin with, for he had provided by-passes at each, so that any one could be thrown out of working without disturbing the daily supply. Moreover, the main was of wrought steel, and therefore not subject to the bursts which might be expected in a cast-iron main, and, in addition, the main was kept under as low a head as possible by regulating the outflowing discharge from Bakers Hill, West Northam, Bulla Bulling, and so on, instead of by means of valves at the respective receiving-reservoirs. When the demand increased largely, it would no doubt be advisable to increase correspondingly the capacity of the service-reservoirs. The advantages of the circular form in service reservoirs, to which Dr. Deacon had drawn attention, were more or less generally acknowledged; and where deemed advisable the circular form had been adopted in the service- and suctionreservoirs of the Coolgardie water-supply, though this was not stated in the Paper. The relative advantages of thick and of thin linings, also raised by Dr. Deacon, must always be one for argument; and, personally, the Author preferred to expend £1 at once, and another £1 after 20 years, rather than £1 158., or even somewhat less, to begin with. The general statement, referred to by Mr. Fitzmaurice, as to the improvement of mortar by the use of a proportion of loam in lieu of the same quantity of sand, was, of course, not based on the Author's opinion, but on the results obtained by Professor Sherman. The tests carried out for the Bulla Bulling reservoir confirmed that writer's results, and the Author would like in this connection to draw attention to the universal Indian practice of adding clay more or less burnt to pure limes. The clay was added in the presence of water, and unquestionably a stronger and more hydraulic mortar was produced than if the place of the clay had been taken by an equal quantity of sand. It was possible that similar good effects were produced by combination of clay with the free lime present in most Portland cements, but extreme care was necessary, as Mr. Hawksley had pointed out.

Professor
Campbell

Correspondence.

Professor J. CAMPBELL BROWN, of Liverpool, agreed that the Brown. theoretical requirements for a good coating for iron pipes were as stated at p. 74. These desiderata, however, were not obtained by the practical means mentioned on p. 80. Not only at Coolgardie, but also elsewhere, Trinidad asphalt was being employed instead of pitch, with gas-tar, for the purpose of coating iron pipes. But

natural asphalt contained only about 14 per cent. of the matter Professor which was really valuable for protecting iron pipes, namely, Campbell Brown. pitch, the remainder being mainly calcium carbonate. This large amount of mineral matter was good for paving purposes, but for coating iron it was not only useless, but positively disadvantageous. It thickened and hardened the pitch with material which did not help to make the pitch adhere to the pipes, and which was gradually dissolved by large quantities of water containing carbonic acid or other substances which acted upon limestone. Accordingly, hard pitch from gasworks would be worth more than seven times as much as natural asphalt for mixing with coal-tar for the inside coating of water-pipes. For the outside it was not of so much importance, because there the coating might be thicker, and the pipe might be embedded in asphalt of considerable thickness with advantage.

Mr. A. FAIRLIE BRUCE observed that the proportion of rainfall Mr. Bruce. actually found to flow off the Coolgardie catchment-area was in accordance with the general experience that, with small precipitation, the percentage received in the reservoir was much less than in the case of a heavy fall, as the former generally consisted more largely of light local showers, and the water was reevaporated or absorbed before it reached the watercourses. No doubt the small percentage collected was also due in part to the gathering-ground being so heavily timbered. Rainfall observations were liable to be very misleading, especially in a large area, unless a considerable number of rain-gauges were distributed over the area so as to give a true average result; as the amount of rainfall was greatly influenced by elevation and exposure. The experience in Bombay had been that the mean rainfall at Tansa Lake, at a height of 362 feet above sea-level, was about 108 inches, at Tulsi Lake, 370 feet high, 104 inches, and at Vehar Lake, 190 feet high, 84 inches. As measured by the rain-gauges, which were situated at or near the dams, about 50 per cent. of the total rainfall reached the Tansa, and almost 65 per cent. the Tulsi and Vehar Lakes; but as much of the gathering-grounds, especially those of the two latter, were much higher than the rain-gauges, it was believed that a deduction of 10 to 20 per cent. should be made from these figures, to obtain correct results. The inaccessibility of the surrounding hills, which were thickly clothed with jungle, prevented rain-gauges from being established on them. The Author was to be congratulated on the general design and water-tightness of the weir, though it appeared to Mr. Bruce that the outlet-works might have been modified with advantage;

Mr. Bruce. for example, by making provision for the fixture of temporary external sluices, to enable the working sluices to be repaired in case of need, as had been done at the Staines reservoirs. The iron stand-pipe in the valve-tower appeared to be somewhat unnecessary. It would have been an improvement had the outer valve-chamber been transformed into a screening-well, as external screens were liable to damage and provided a very small area. The precautions taken to protect the toe of the dam from the effects of scour by water flowing over it were wise, though the impact might have been further reduced had the outer face been stepped or rock-faced. At Tansa, where of course the body of water flowing over the portion of the dam used as a waste weir was much larger, nullahs, 20 to 30 feet deep and 50 to 100 feet wide, had been cut by the water, and sooner or later protection works would become necessary; as already at one or two points the erosion had approached nearer to the foundations than was altogether desirable. The locking-bar pipe adopted seemed to have been in every respect the most satisfactory form of pipe that could have been used, as, in addition to avoiding the loss of strength due to riveting, considerable longitudinal stiffness was imparted by the bars. It would be interesting to know if the coating was proving durable, and if a perfectly smooth internal surface had been obtained. In America where several asphaltic compositions had been employed for coating steel pipes, they were usually baked in an oven to harden and improve the surface. The Author did not mention whether yarn had been used in making the joints, or if not, what method had been adopted to prevent lead from running into the pipes. On the Glasgow waterworks, where the 48-inch pipes were jointed with collars of somewhat similar design, this was effected by means of internal rings, which were removed after the joint was run, when it was staved internally as well as externally. The reduction of diameter at the valves might have been carried even further with considerable advantage. Mr. Bruce had for some years employed 24-inch valves for 48-inch pipes, and 18-inch valves for 32-inch pipes, small valves being cheaper in first cost and much more easily manipulated, repaired, or replaced, than those of the full diameter of the pipe. Kutter's formula was quite unnecessarily complicated for calculating the discharge of pipes, that of Darcy being much simpler and more logical. Applying it to this case with a coefficient of friction of 0.00345, the results obtained by the Author would appear to be about 18 per cent. less than they ought to have been for new and well-coated pipes; and it would be interesting to know if there was anything

in the roughness of the coating, or any friction due to the projection Mr. Bruce. of the bar into the cross section of the pipe, which would account for this. The following were the results of some experiments made by Mr. Bruce on the discharge of large pipes:

The covering of the pipes with earth would appear to have been a wise step, though the banks would demand constant attention to prevent them from being washed or blown away. Experience in Bombay had shown that where pipes were laid below ground in dry earth, free from salt or organic impurity, they had remained for nearly 50 years practically unaffected by corrosion; whereas others laid above ground had been more or less attacked in a quarter of that period. They also required frequent caulking, owing to the expansion and contraction due to variations of temperature. Concrete lining on slopes, such as that employed in the service- and regulating-reservoirs, was usually found to give trouble, though it was possible that the system of reinforcement adopted might suffice to prevent the tendency to crack owing to slight subsidence, which generally occurred sooner or later. Had the floor been asphalted the thickness of concrete might have been safely reduced to about 6 inches, without danger of leakage. Mr. FOSTER CROWELL remarked that the Paper, besides being Mr. Crowell. an instructive record of excellent work, contained useful suggestions of general bearing. For example, the facts cited in regard to the faults discovered in the rock bed at the site of the Helena weir showed the meagre value of preliminary borings and trialshafts as indicators of the conditions of any rock formation, with reference to its continuity and reliability for the purposes of a dam. This was a not infrequent experience, and had been marked by many striking examples, a prominent one being the New Croton dam, of the New York City waterworks, recently completed, where extensive and careful preliminary explorations had been made, but where great discrepancies