Mr. Crowell. had been disclosed later by the complete excavations.1 It might be set down almost as an axiom that the suitability of a particular rock formation for the foundation of a dam could not be known definitely until the final excavations were actually made: nevertheless, it was customary for engineers to proceed with the design, make the preliminary estimate of cost, and award the contract, as if the borings and trial-shafts were quite sufficient, trusting to luck or depending on judgment as to whether the indications were comprehensive. If afterwards serious geological faults were disclosed by the excavation, a change or abandonment of site might become necessary, or at least desirable; but usually in such cases one of two courses was adopted, either going deeper-involving increased quantities and cost, and perhaps disclosing new difficulties-or resorting to some more or less unsatisfactory artifice with attendant risks. Guided by past experience, he had adopted some time ago a rule to assign only negative or tentative values to preliminary borings; that was to say, the absence of rock would be a positive indication of unsuitability, but no other indications would be considered positive until corroborated. Carrying out this principle in preparing plans for a masonry power-dam at St. Joseph's, N.Y., in 1903, subsequently built under his direction, he had decided to ascertain actual conditions by uncovering a large part of the proposed site down to bed-rock at the owner's expense, before completing the plans and specifications or inviting tenders. There had been special reasons for precautions here, owing to the fact that the proposed structure was to take the place of a former dam which had failed because of defective foundations. Fortunately, favourable conditions had been found; but even if it had turned out otherwise, the extra expense entailed in that contingency, regarded as insurance charges against a much greater loss, would have been amply justified, and he would advocate a similar course in all important cases. It was of interest to note the growing co-ordination of recognized engineering methods on modern works, throughout the world. It did not often happen, of course, that conditions were precisely the same in different countries; but, especially during the past few years, it had come to pass that, given like general conditions, the treatment resorted to by engineers was apt to be carried out on closely similar general lines. Even in the details, and in the technical terminology employed 1 C. S. Gowen, "The Foundations of the New Croton Dam." Transactions American Society of Civil Engineers, vol. xliii. p. 469. in describing results, the same was true; so that in reading and Mr. Crowell. studying in America this Paper, written of work carried out in Australia-which, to Americans, was one of the uttermost parts of the earth-no glossary was required in order to arrive at a full understanding; while the work described was excellent and up-to-date, if, indeed, it was not distinctly in advance in several respects. The pipe-line was noteworthy, in respect of both the use of locking-bar pipe, and the very ingenious and effective caulking-machine; its great length and high working-pressures gave it rank among notable metal conduits; and its very fine results, as shown in the Paper, were matters of congratulation to designer and maker alike. The discharge-capacity of locking-bar pipe compared with riveted pipe of equal diameter, as shown by the tests, was a matter of great interest and importance, and it was to be hoped that the Author would supplement these tests from time to time with others, to determine the effect of use on the degree of roughness and on the capacity. It would be agreed by all interested in hydraulic matters that there was hardly any subject connected with the science of water-supply that was so greatly in need of reliable data as was the question of flow in pipes; and the Coolgardie conduit, with its arrangement of receiving- and regulating-tanks, afforded peculiar facilities for making exact and reliable determinations. There were a number of other points, such as durability and progressive leakage, permanence of pipe-coating, etc., which periodical tests would bring out. So far as Mr. Crowell knew, it was not usual to pay great attention to the procurement of such data; but they would prove of great value if systematically collected and published. Mr. WILLIAM FAIRLEY thought that the description of the Mr. Fairley. pumping-machinery might have been enlarged with advantage, considering the very important part it played in the success of the works. It would appear that, from the initiation of the scheme, considerable uncertainty was entertained as to exactly how much water would be taken during the first few years after the completion of the works; the quantity having been put as low as 1 million gallons per day. In these circumstances, while it might have been advisable to construct the conduit to the dimensions necessary to deliver the full quantity provided for in the scheme, the necessity for providing the pumping-machinery on a similar scale was not apparent from the information given in the Paper. In the first four stations there were provided three units, each approximately of 300 HP. and equal to one-half of the required capacity, with one of the units, or one-third of the total power, in reserve. Mr. Fairley. Boiler-plant was provided on a similar scale, and this arrangement appeared to give a safe reserve and to be particularly suited to the case. In the remaining four stations, however, only two units were provided, each equal to the required full capacity, one of the units being spare, with boiler-plant similar. The reason given for this variation was that economy was effected by having the whole of the steam ends of the engines to the same patterns, thereby reducing initial outlay and the number of spare parts to be kept in stock but in a scheme of such magnitude, where about twenty sets of engines were required, sufficient saving by having one size of engine instead of two or three different sizes, was not apparent. Certainly from the engine-builder's point of view it was convenient, and might mean not only that he realized some little economy in constructing in the shops, but also that power considerably above actual requirements had to be installed in four stations. The advantage of having all spare parts to the same standard for eight stations, spread over 300 miles of country, did not appear to counterbalance the serious expense in providing additional engineand boiler-power. It did appear that if three units, as in the first four stations, had been provided in the remaining four, giving each station one-third reserve, a very considerable saving would have been effected; and in any case the smaller units for pumping and for steam-generation would have lent themselves more readily to economical working in the early years, should the demand be less than originally estimated-which Mr. Fairley understood was the case in this instance. Instead of 6,068 HP., as shown provided in the Paper, 5,463 HP. would appear to be sufficient if two standard sizes had been adopted instead of one. The difference, at £48 per HP., amounted to about £29,000. The practice of putting down two units, each equal to half the maximum demand, and adding the reserve third as soon as required, would appear to be more economical, and might have been applied in this instance; there would have been considerable saving in the capital cost and also no difficulty in adding further pumping-power as required in the future. On p. 89 the Author mentioned that the feed-water was passed through a heater placed on the exhaust-pipe leading to the condenser. Mr. Fairley would be glad to know if experience had shown any gain from this arrangement. Also, had the Author found any difficulty in keeping each pumping group doing its exact amount of work, due to the tendency in this type of pumpingengine to run short stroke? Mr. Folwell. Mr. A. PRESCOTT FOLWELL, of Easton, Pa., remarked that, while it could hardly be questioned that the yield of each catchment-area was a law unto itself, it was also true that watersheds similarly Mr. Folwell. located had many characteristics in common; and in view of the apparent similarity in several respects between the Helena basin. and many of those in the south-western part of the United States, a brief statement concerning these latter might be of interest. Along the Pacific coast of Southern California ran a range of mountains, east of which lay about 80,000 square miles of country, between 4,000 feet and 10,000 feet above sea-level, which had a mean annual rainfall varying between 2.8 inches in the south and 7 inches in the north, with an occasional year of practically no rain at all. As an illustration of the yield in this region, in 1901 one area of 4,000 square miles yielded 0.185 inch, and two others, of about 23,500 square miles each, yielded 0.246 and 0.2785 inch, respectively. In the mountainous region to the west the mean rainfall varied between 5 inches and 60 inches per annum, and several reservoirs were located there. Typical of many of these was the Sweetwater, which was notable in possessing fairly complete records of rainfall, evaporation and yield on the drainage-area, dating back to 1888. The rainfall and the ratio of the yield thereto were given in the following Table, the years extending from September to August inclusive: RAINFALL AND YIELD ON THE SWEETWATER CATCHMENT-BASIN. Practically all the rain fell between September and April, or during about the same length of time as the rainy season at Mundaring; and it was seen to be approximately similar to the rainfall there in both quantity and variations. The percentage running off, however, averaged almost five times as much as at Mundaring. In Fig. 35 the rainfall and the percentage of yield for each of these two watersheds was plotted, and a remark RATIO OF 4% 3% 2% 1% 3% the able similarity between two was evident, both showing what the Paper referred to, namely, that the ratio of yield to rainfall was even more variable than the rainfall itself; also, that extremes of rainfall were likely to be accompanied by still greater extremes in the yield-ratio, the ratio-curve always moving in the same direction as the rainfall-curve. At Mundaring, however, this ratio had not reached zero, and it was 0.5% to be hoped that it might never do so; but in all the dry regions of the United States such a condition was likely to occur more or less infrequently. As on the Helena, the high percentdue to ages of yield were heavy storms, during which onefourth to two-thirds of the entire season's rainfall fell in 2 or 3 days. The last seven seasons on the Sweetwater conveyed a 0.5% lesson which would not soon be forgotten in that country, for in only one of these had the yield equalled the evaporation from the reservoir (annually 15 to 20 per cent. of the reservoir's capacity), which had consequently been almost entirely empty for 4 years, although having a capacity of about three times the consumption. It was possible that the apparent similarity between these two cases was but superficial; but if not, the storage allowance at Mundaring, equal to 2 years' consumption, was not 2% 1% 24% 20% +16% 12% 8% 4% |