Democrat and Chronicle from Rochester, New York • Page 10

KOCIIESTEIt DEMOCRAT AND CHEOXICLE, SUNDAY, iTAXUARY 27, 1895. 10 "inside" courses, which fitted closely Into and the pipe work separately; the latter atmosphere as soon as the water nreasnr. i removed, since the tension of the m.rti.., only sufficient to balance the wJIgJf'V1 delivery of the conduit, by suitably throttling for a portion of the day a few of the stop-valves in the principal mains at the foot of the hill on which the distributing reservoir is situated. This plan was accordingly adopted in June, 1W, and was retained with some modifications until Au- SntVhouid be remarked that up to June. 1890 it had teen taken for granted that no appreciable diminution in the discharging capacity of the old conduit had occurred, and that it was stil continuing to deliver about 9.000,000 gallons per day.

At that time, however, a close gauging of its now into Rush Reservoir was made, which revealed the fact that its actual delivery was then only about gallons per day. The discrepancy between these two discharges could not be accounted for by leakage or abstraction of water from any portion of the conduit, and hence it was evident that its efficiency had become permanently impaired. Meanwhile, Chief Engineer Tubbs became ill and resigned his position at the end of July, 1890. A few weeks later the writer, v. ho had severed his connection with the Water Works Department some years previously and was engaged elsewhere, was appointed to take charge of the "fnthls state of affairs, the matter of obtaining a temporary additional supply of potable water was considered of chief Importance.

Some months previously an old artesian well in the town of Gates, about one mile west of the city, gave promise of furinshing a supply of 1,000,000 gallons per day, and contracts had been made for the rental of this well, and also for the delivery of a suitable pumping engine and the necessary quantity of pipe, for the purpose of forcing this water directly into the nearest city distributing pipe. The machinery had. however, not vet arrived, and as some doubts regarding the permanency of the flow from the well had Keen entertained, preparations were made to test its capacity in a time of drought. The result of this test was very discouraging, and as the owner of the well refused to allow any new wells to be drilled anywhere on his land, also because the geological characteristics of the region were unfavorable to the assumption that the same vein of water could be tapped by other wells at some distance from the original one, it was deemed prudent to investigate other sources before incurring the large expense involved bv the construction of the above-described temporary works, especially as it was very doubtful whether the well would yield even 500,000 gallons per day during dry periods. To complete our outline of this part of the subject, it may also be mentioned that none of the various propositions for tern porary relief, which were advance after the failure of the Gates well, met with ap- 1 at T--h IMS whpn an agree- Hrput pure was also reoulred In the hand ling and transportation of the pipes, both by railway and by wagon, to avoid damaging the coating.

To this end, all chains and bolsters or saddles which came In contact with the pipe were covered with old rubber hose, canvas or carpeting, and the site where each pipe was deposited, preparatory to excavating the trench, was thoroughly cleared of stnnes. In spite of these precautions, the abrasion of the coating in spots whs unavoidable, not only from the processes of transportation, but also from the subsequent processes of riveting the pipes together. It may also be mentioned that duriug the latter work, canvas coverings or mats were placed on the top of the pipe outside, and on the bottom inside, wherever the workmen were engaged; and all who entered the pipe for any reasons were required to use soft felt or rubber overshoes. As it was essential that every defective place In the coating should be repaired as well as possible, the use of a suitable asphaltic paint was necessary, and in the absence of anything better, a preparation called B. Paint" was selected for the purpose.

This material consists of refined asphalt dissolved In bisulphide of car-lMn; and as the latter Is a very volatile liquid. Its rapid evaporation soon leaves a thick coating of hard asphalt on the metal. The offensive odor of this liquid, however, is a serious objection to its use on the interior of the pipe, as the vapor is both suffocating anil inflammable; hence it was necessary to Insure the thorough ventilation of the conduit during its instruction. For this purpose, as well as to afford future ingress for examination and repair, manholes with tightly fitting covers were riveted upon the pipe at Intervals of about l.OtK) feet. The covers of two or three consecutive openings were removed while the interior of the conduit was being examined, and newly driven rivets nnd abrasions in the coating were being painted, and a strong current of air was forced through these portions of the pipe bv means of a suitable fan or blower.

In this manner the performance of the necessary Inside painting work and its thorough Inspection was made endurable. LAYING THE CONDUIT. Description of the Manner of Placing the Pipe in the Trenches. It should also be stated that after the pipes hail been delivered along the, route of the conduit, two lengths were riveted together upon the surface of the ground, thus forming a section about 4.5 feet In length, ami weighing from to pounds, according to the class. These sections were then rolled in succession by a large number of men upon suitable timbers placed across the trench, where one or two strong derricks raised them clear of the timbers ami lowered them safelv to the bottom of the excavation, one end of a section was then carefully fitted into the pipe which had previously been laid, and was temporarily fastened thereto with a number of bolts paused through the rive holes.

The exact adjustment of the nectfon to Its proper alignment and grade was then done, and after the rivet holes had all been made to coincide, either by reaming or a slight turning of the pipe, the joint was ready for the riveters. Meanwhile the trench had been excavated truly to line and grade, and wherever the poll "permit ted. its bottom was shaped to fit the convex exterior of the pipe. All large stones nnd boulders were carefully removed therefrom, and in rock excavation the trench was made somewhat deeper and then re-tilled to grade with earth. Where the bottom could not be shaped as aforesaid, woollen saddle blocks were placed at frequent intervals and brought to grade by temping the soil underneath.

At the joints, the trench was enlarged somewhat, so hs to admit of Inserting the hot riv-ts from the bottom ami sides, the upsetting and hammering of these rivets being done on the inside of the pipe. The remainder of the rivets were passed into the pipe through a small hole cut in its top. and were upset on the outside whilst being beld in place with a lever on the inside. On the completion of the riveting, the pass-hole v.as closed with a brass plug firmly screwed Into place. These holes also served to assist in ventilating the pipe, as well as to facilitate the Introduction of fresh supplies of paint to the workmen It-side.

In refilling the trench, only the finest, selected earth materia! was deposited under and around the pipe up to a level of six inches above Its top. and was thoroughly rammed in thin layers as it was being thrown so as to give the pipe a solid bearing for its entire length. Somewhat less care was taken with the tilling of the remainder of the trench, although the material was also impacted bv ramming where required. For the greater part of the route. It became necessary to partially refill the trench soon after the pipe bad been riveted together, in order to prevent the latter from floating In case that the excavation should accidentally be filled with water: but wherever practicable a space was left open around every field-made joint, in order that any leakage might easily be disi-overed and remedied on the application of the water-).

ressure test to the conduit. On the other hand. In places where no appreciable inflow of water could nr. the upper part of the pipe was left entirely exposed until after said test, thereby obtaining an opportunity for observing the qualify of the shop-made seams. I'pon the completion of lengths of a mile or more of the conduit, each such section was in Its turn thoroughly denned out and ex amined to see that all' defects in the coating hail been made rood with paint, wnereupon thA ends were tight Iv closed up.

either tem porarily with a steel boiler head or permanently 'with a stop valve, and the pipe filled with water from a force pump. The water pressure was gradually ralsinl until it became nhotif nt) mt cent, creater than that to which the conduit would afterward be subjected in its normal operation; and while under this pressure tne worn was cioei.r examined for leakage. By this means all defects quickly manifested themselves and were thereupon promptly repaired. It may also be mentioned that the force pump was usually stationed at the beginning of the ripe conduit near the village of Hemlock t.ake, nnd hence most of the sections above Itush reservoir were thus piibjected repeated ly to this severe Test. For the portion of the conduit from Kusn to Jit.

Hope reservoir, the force pump was not needed, as the static head from the said beginning of the work was amply sutlicieut for testing pur poses. STAND-PIPES AND AIR-VALVES. Novel Devices Along the Line of the Conduit Their Purpose; A comparatively novel feature of the pipe conduit is the provision of a device, midway in the long stretch of 17.. miles from Its "commencement to Hush reservoir, whereby the pressures lu the upper half will, under ordinary conditions of operating the works, alwars 'be limited to those due to the hydraulic gradient: while the maximum pressures In the lower half, which will result If the inlet valve to said reservoir be closed, will be about pounds per square inch less than those due to the entire available head. The device consists simply of a liort stand-pipe, rising from the conduit on the top of a hill up to the hydraulic grade line for Hush reservoir, ami communicating with an adjacent pipe which Is continued down the hillside to Ihe creek.

With the flow properly regulated by the gates at the lake, and free discharge into said reservoir, the water will tint rise In the stand-pipe quite tip to the level of the overllow, and hence no wnste will ensue; but If the discharge at Uush Is stopped tn some emergency, the overflow will at once come into action and carry the entire flow safely Into the creek, thereby the head on the upper half of the line from acting on the lower half. The resulting loss of water from the lake will be trifling, as word can soon lie sent to the latter place to stop the supply. On the other hand, the benefit derived from the device Is the ability to use thinner plates for the pipe over a long distance, and thereby materially diminish the cost of the work. At every summit In th conduit line, an air-valve of peculiar construction has been placed. The function of these appurtenances is Iwith to allord a facility for letting out the air which may be confined in the pipe before It is tilled with water, nnd to allow air In abundance to enter automatically.

In case that water Is drawn out from blow-offs or escapes from rupture In the pipe The investigations for determining the pr per size of such a fixture at any particular summit are extremely Interesting, but lack of space prevents their Introduction here. It will therefore suffice to say that three different sizes, affording clear openings of IS, 4 and 0 Inches in diameter, were used In the work. The devices consist, respectively, of one. two and four similar bronze valves, suspended from springs nnd attached to a su'table cast-iron chamber, which is bolted to horizontal stop-valve, and the latter to a Hanged casting riveted upon the plp. The stop-vnlve is left wide open, but If the nlr-vnlve gets out of order, the former may be closed and the repair made without shutting elf the conduit.

As It Is necessary to keep these fixtures free from frost, earth and mischief, they are Inclosed in an iron case deeply covered with earth, and from which a capacious stand-pipe protects alsuit three feet above the surface of the ground. The top of this Ktnnd-nipe is covered with an Iron hood. In which there Is a smnll opening, also covered with locked Hup. to receive the key whereby one if the cluster of lr-vnlves may be pressed down and opened. For the admission of air.

the valves are all opened automatically by the pressure of the from nearly the whole of the northwestern quarter of Monroe County. This area is relatively thickly populated, and contains a large number of villages and hamlets. In view of all these circumstances, the nearest available location for an intake on the lake shore is at a point about 8.5 miles west of the mouth of the river, and 15 miles in a direct line northwesterly from Mt. Hope Reservoir. By the shortest practicable route, without crossing the populous districts of the city, this distance would become about 17 miles.

Briefly summarized, the various reports and estimates showed that the long force main, costly intake works and pumping engines, and the capitalization of the annual expenses for lifting the water to so high an elevation, would render the Lake Ontario project far more expensive than either of the other two: and that while a conduit from Conesus Lake might cost a little less than one from Hemlock Lake, yet the expense of protecting the water from pollution, by introducing and maintaining adequate sanitary arrangements on the large and relatively populous watershed of conesus Lake, would probably more than make up the difference. Various other elements also entered into the choice between these sources, but after due consideration, the preference was finally given to Hemlock Lake by all the municipal authorities, on June 16, 1891. it roa Pending this definite action, a bill was troduced in the Legislature early in 18.41, whereby the city would be allowed to take an additional water su-jply from either Hemlock or Conesus Lake, and to bond itself for the necessary costs in the sum of $1 750,000. For some reason, however, the measure was opposed, and the Legislature adjourned before the bill could be passed. As a large amount of engineering work, both in the field and the office was necessary before the details of any of the general plans could be prepared, and no funds for the purpose being available, the Common Council appropriated an adequate sum on June 26, 1891.

ADOPTION OF PLANS. Detailed Description of the Construction of the New Conduit. Soon afterward, two surveying parties commenced operations and were kept busy during the remainder of ths year examining various routes and preparing topographical maps thereof. From the data thus acquired careful studies and comparative estimates of the most feasible routes were then made, and a definite location of the conduit established. The previous hydrolog-ical investigations were also thoroughly revised for the purpose of determining the storage volume needed in the lake during a.

cycle of years of least observed rainfall, and from' this the necessary depth of the upper end of the conduit below the existing ordinary low-water level of the lake was fixed. Nearly a year was occupied in this work and in the preparation of the drawings and specifications. A new bill similar to the one above mentioned was introduced in the Legislature early in 1892 and succeeded in becoming a law on April 20, 1892. Work on the new conduit plans was therefore continued energetically, and proceedings were at once instituted to obtain the necessary rights of way in those localities where the route had been fixed. Much delay was here encountered, owing to the organized opposition of a number of landowners to grant such rights except at exorbitant rates, and advantage was taken of the time during which these matters were in court to have all the plans and specifications carefully reviewed by Mr.

A. Fteley. Chief Engineer of the Croton Aqueduct Commission, who was retained as Consulting Engineer by the Executive Board, while the general provisions of the forms of contract were to be criticised by Mr. C. D.

Kiehel, then City Attorney. The plans for the new conduit as finally adopted contemplated the construction of a brick conduit, of horse-shoe shaped cross-section, about six feet in diameter, on a grade of one in 4,000 from the northeastern corner of Hemlock Lake to a point on the east bank of the creek about 12.i feet northerly. Of this length, T.fxiO feet was in tunnel worked from several shafts through the clay and shale rock which un-deriies the region, while the remainder was in open excavation. A large and deep gate and screen house was to be built at the shore, from which a steel intake pipe, 5 feet in diameter, and l.fioO feet long, was to lie extended into the lake to a point where the water was about 35 feet deep. The pipe was to be laid in a channel dredged to a depth of about 19 feet below the existing ordinary low-water level of the lake urtil this depth of water was reached, after which it would be laid directly on the bottom and terminate in a submerged timber crib resting on the bottom.

At its beginning in the gate house, the invert of the brick conduit was to be about 17 feet below the aforesaid low-water level; and at the northerly terminus it was to be a little above the average water surface In the creek, in order to provide for natural drainage in case of cleaning repair. To prevent the masonry of the conduit from being injured by the water pressure which would ensue from improper manipulation of the works, an overflow chamber was to be constructed at the north end, whereby the water surface in the conduit would effectually be prevented from reaching the top of the arch. From the said overflow chamber onward to the storage and distributing reservoirs, the water was to be conducted in either a 30-inch cast-iron, or a 38-inch steel pipe comluit, laid on a continuous hydraulic grade to its terminus at the latter reservoir. This grade was about one in 570 for the total distance of nearly 110,000 feet, or 26.5 miles. The route was along the east side of the relatively narrow outlet valley with steep banks for a distance of about 2.0 miles to Richmond Mills; thence northerly for about 3.5 miles across the undulating territory west of Honeoye Creek to a point where the valley again becomes narrow with high banks; thence still northerly a similar distance through the said valley, in which the creek meanders from side to side, the pipe crossing and re-crossing the same nine times to avoid tunneling and location on treacherous banks; thenar! rising on the east side of the valley to the elevated plateau below the village of West Bloomfield.

and continuing northerly along this plateau and across the deep depression called Surrine Hollow, for a distance of 5.35 miles; Jhence northwesterly along a highway and across fields for 3.41 miles in a direct line to Rush Reservoir; thence northerly along the route of the old conduit and through the Henrietta road for 4.81 miles to a point where the prolongation of Pinnacle Avenue intersects said road thence along said prolongation to and through Pinnacle Avenue for 3.31 miles to a point 4j feet north of Elmwood Avenue: and thence northwesterly across private and park lands lands for 0.69 mile to the east end of Mt. Hope Reservoir. New gate houses and inlet works were also projected at both reservoirs. It should be stated that the Pinnacle Avenue route was chosen, because it was foreseen that a new and large distributing reservoir would soon be required; and as the most available site for such a structure is on the elevation known as Cobbs Hill, at the southeastern corner of the city, while Pinnacle Avenue is nearly midway be-ween this site and Mt. Hope Reservoir, it follows that the said location of the pipe conduit would best serve both places.

A few words, also, as to the reason for providing so large a masonry conduit and intake works at Hemlock Lake may be given. The enormous difficulties with quicksand which were encountered in the construction of the old conduit in this locality, and the discovery of extensive deposits of this material by preliminary borings along several contemplated routes for the new conduit indicated that the costs of construction for the lter would be so great, that the addition of 1.5 or 2.0 feet to the diameter needed for a discharge of 15,000.000 gallons per day, would not make a serious increase in the amount; and as this additional, size would give the conduit a oatvacity for delivering easily about 32,000.000 gallons per day, which was the computed average yield of the entire watershed of Hemlock and Canadice Lakes, it was deemed prudent to adopt the larger dimensions in order that no further outlay in this locality would arise, if a second additional water supply should be required in the future. The same considerations are also applicable with even greater force in the case of tunnel work, owing to the necessity of having room enough for the workmen to perform their labor properly and quickly. LETTING THE CONTRACTS. Care Taken That the Work Should Done in a Proper Manner.

be Considerable time was occupied in the work of revising and printing the several voluminous documents constituting the forms of notice of letting, proposal, contract, specifications and bonds. Early in November the matter was finally submitted to the Common Council for adoption, which wa3 formally done on November 15th, and the entire work was then extensively advertised for letting to take place on December 23, 1S92. Bids were requested from contractors for the masonry conduit work ind alternated witn tne iuiuiw. diameter of every Inside course was made exactly 38 inches throughout, and the net length of both kinds of courses, estimated from the centers of the rivet holes at the ends, ranged from 6.83 to 6.i leet, according to the thickness of the plates. Each course or section, moreover, was made of a single sheet, of such width as to form the required cylinder with ample allowance for overlap.

Four classes of pipe, formed of three air-ferent thicknesses of plate, were manufactured and arranged according to the water pressure to which they would be subjected in the line of the conduit. For pressures due to static heads of 120 feet or less, the pipe was made of plates Vi-lnch thicK, with the longitudinal or straight seams single riveted; for heads ranging from 120 to 153 feet, the same thickness was used, but the straight seams were double riveted; for heads between 153 and 199 feet, the thickness was 5-16-lnch, and for heads between 199 and 263 feet, the thickness was -inch, with the straight seams double riveted in both classes. All of the circular or round seams were single riveted, except where two different classes of pine were joined together, in which case these seams in the thinner pipe, including the junction seam, were double riveted for a distance of about 200 feet, ine object of this provision was to make the efficiency of the double riveted round seams in the thinner pipe approximately equal to that of the single riveted round seams in the thicker pipe, as it was computed that in the distance mentioned, the friction of the earth against the pipe after the trench is properly refilled, would be sufficient to balance the resultant of the longitudinal forces produced in the two classes of pipe by the assumed extreme variation of temperature in the finished conduit, which was here taken at 45 degree F. Each class of pipe is therefore regarded as being anchored in the soil, except for the said distance at the junctions. The pipe was generally furnished lengths of about 27 1-3 feet, consisting of four courses riveted together in the shop with powerful machines, and having all seams thoroughly calked on both the inside and the outside.

Changes of direction, either in alignment or grade, were made by slightly beveling, at one end only, as many courses as were needed to make the required curve or deflection. These beveled pieces were then power-riveted in the shop, either to each other or to square-ended courses, according to whether the curve was of short or long radius. In the former case it was found expedient to rivet only two courses together in the shop, on account of difficulties in the transportation of longer pieces; but for easy curves requiring only one beveled course to three square-ended ones, four courses were riveted together the same as for straight pipe. To avoid multiplicity of patterns, a standard bevel, of 2 degrees and 43'4 minutes, corresponding to an offset of 1.8(in inches in the diameter of 38 inches, was adopted and 'ar as lKJSsible throughout the work. ith this deflection it was still practicable to make a good circular seam without flanging or hammering over the ends of the courses If a somewhat smaller angle was needed occasionally, a special pattern therefor was made; and deflections of less than one-third degree were easily accomplished by slightly reaming out the rivet holes on opposite sides in the round seams.

The above-mentioned standard bevel on the end of every fourth course forms a regular polygon which fits closely what railroad engineers designate a 10-degree curve; applied on the end of every second course, it corresponds to a 20-degree 8-minute curve, and on the end of every course it gives a 41-decree 9i-minute curve. These three curves may also be defined in terms of their radii, which are :,.3.7, 28b.l and 142.2 feet respectively; and all changes in direction' and grade of the pipe conduit were made in accordance therewith. It may also be remarked that the work was in general so arranged as to avoid making a change of grade In a curve, although In a few instances such a combination could not be obviated without Incurring large additional exper.ES for excavation. PREPARING THE PIPE. Preservation of the Metal Was the End Always in View.

As soon as pratlcable after the riveting and calking on a pipe was finished, it was thoroughly cleaned, heated in an- oven to about 30 degrees and then placed in a large tank containing the hot coating mixture where it was allowed to remain for such time as would suffice for the tempera-cure of the metal to become the same as that of the mixture. The pie was then removed from the bath and the surplus coating substance allowed to drip off. after which its treatment by the two different manufacturers varied very materially. At the East Jersey Pipe Works, a purely asphaltic coating was used which became hard when cold; and hence after having been dipped the pipes were simply placed uifMa r.iiixide. where the material was Weft to harden.

At the Rochester Bridge Iron Works, on the other hand, a spe cies of japan coating was useu, mi ii inquired the pipe to be placed vertically in a large brick oven after dipping, and baked therein at a high temperature for about ten hours, on an average. Much attention was given to the matter of preparing and applying a coating mixture which should be durable, hard, tough and adhesive to the smooth surface of the steel plates at all ordinary temperatures. Numerous experiments with different kinds of asphalt and other substances had been made by the writer, at intervals during several years before the new conduit was projected, in the hope of finding a compound which would meet all of these requirements; but as they were generally more or less unsatisfactory In fulfilling the last-named condition, it was deemed expedient, after the decision to use steel pipe in the new work was reached, to seek counsel from some chemist who had made the preparation of paints, varnishes and lacquers a special study. In response to various inquiries, the writer was referred to Prof A. H.

Sabin. of the well-known firm of Edward Smith varnish manufacturers. At New York, ami finally succeeded In persuading him to undertake the necessary investigations. Much time, however, was required for this purpose; and as the East Jersey Pipe Works were soon ready to commence making pipe, the proprietors were directed to use meanwhile the same asphalt coating which had been found to be very successful on numerous conduits in California, as well as on the large steel conduit recently completed for the water supply of Newark. N.

J. This material hail been highly commended by several distinguished chemists, and it was generally regarded as superior to any other preparation of asphalt which had yet been devised. effort was made to obtain materials of the best quality for the purpose from California, also to prepare and apply the coating in the most approved manner. The results at first appeared to be very gratifying, but after the pipes had been exposed to the air for several weeks, it was noticed that the high gloss of this coating gradually disappeared, thus leading to the inference that the substance was undergoing a slow change. In water, on the other hand, no alteration was detected, and it was therefore believed that the coating would prove serviceable after the pipes were once filled.

It also failed in adhering tightly to the metal, and the contractors were subjected to great expense In painting with costly material large areas from which the coating had scaled off during transportation and handling. Directions were then given to substitute for the California asphalt, a mixture of refined Trinidad asphalt and the best grade of coal-tar, thereby producing the same coating as had been applied to the old conduit, and which hns proved to be durable at all events. Although this mixture was. on the whule, more satisfactory than the other one. it nevertheless left much to be desired in the way of toughness and adhesion to the steel.

Meanwhile the experiments of Prof. Sabin In the direction of obtaining a better practicable coating, had resulted in the evolution of a japanlng process which, from both theoretical ami practical considerations, gave "promise of conspicuous success 1'ieces of small pipe, which he had coated in his laboratory, were subjected to numerous tests for determining the durability of the coating; and as it withstood all these tests perfectly, and. moreover, adhered remarkably well to the metal. Its use for the remainder of tne conduit was de-decided upon. At this time, however, the Knst Jersey Pipe Works tuid nearly coin-nleted their portion of tlie work, while the Hochester Bridge nnd Iron Works wns just commencing operation on the pipe.

The proprietor of the latter was accordingly requested to Investigate the Sabln process thoroughly, nnd to adopt It If no valid objections could be raised ng.iliist It by the expert chemists he was to employ for the purpose. As no such objections were forthcoming and nothing better was available, ho promptly constructed the necessary expensive plant for carrying out the process under Prof. Knbin'H direction, and coated all of his pipe in this manner. So far a can now be determined, this coating Is superior in all respects to those which were applied to the first half of the conduit, and It Is hoped that the crucial test of time will fully confirm this opinion. STORY OF A BIG JOB Description of the Construction of the New Water Works Conduit.

DETAILS OF THE WORK Facts and Figures Presented by Emil Kuichling, the Engineer in Charge-Full Text of the Paper Read at the Chamber of Commerce. Following is printed the full text of the exhaustive aper on the Rochester water works system that was read at the Chamber of Commerce rooms last Monday night by Engineer Emil Kuichling. The paper is interesting as well as complete, the great maws of details regarding the work boing presented in attractive form. The meeting at which Mr. Kuichliug was heard was held under the auspices of the Academy of Science.

Following is the paper: entering upon the subject of this paper, it may perhaps be expedient to give a short account of the old water works of the city, and the circumstances which led to the necessity for an additional supply. The question of providing- a public water supply for the city had been considered for some vears prior to 1860, at which time the population was about 48.000; but it never assumed definite form until that year, when a private corporation obtained a franchise for bringing water by gravity from Hemlock LAke, which is situated in a hilly district about 29 miles south of the city. This corporation, however, failed to accomplish its purpose after spending several hundred thousand dollars in the attempt; and as there was no evidence that the undertaking would ever be properly completed, the city finally resolved to build an entirely independent system of water works. A Hoard of Water Commissioners was accordingly created in April, IS. 2, and on November lrith, of the same year, this board reported in favor of a dual system of sup-plv, selecting Hemlock Lake as the most expedient and desirable source for obtaining by gravity an abundance of pure, soft water for domestic and general uses, and supplementing the lack of efficiency of the gravity work3 for fire purposes in the central and manufacturing districts of the city, by a separate direct pressure pumping system supplied with hard and impure water from the Genesee River, which flows northerly through the town.

The final plans for this project contemplated an ultimate capacity of about gallons per day from each of the two svstems. to be distributed through about 40 miles of pipe in the city streets; they also provided for the construction of two reservoirs on the line of the gravity conduit, one on the southern boundary of the citv for distributing purposes, and the other about nine miles south of the city for storage purposes. The total expenditure lor the work was limited to $3,000,000. The general plan thus outlined was duly approved bv the municipal authorities, and early In 1873 the work was advertised for letting Meanwhile the necessary details of the undertaking had been elaborated under the direction of the Chief Engineer, Mr. Nelson Tubbs, and construction was commenced soon after the contract was awarded.

As it was anticipated that the long gravity conduit would require nearly three seasons to finish, the direct pressure system was comnleted first and put in full operation in February, 1S74, with 7.32 miles of distributing mains. The gravity works, with he two reservoirs, a compound conduit consisting of 9.75 miles of SB-inch riveted wrought-iron pipe and 18.50 miles of 24-inch pipe, mainly of cast-iron, and 51.52 miles of distributing pipe, were finished early 1870 It may be remarked that when these works were tne cuy nan a 1.1u-lation of about 70,000, and an area of 4,840 acres with lb6 miles of streets and alleys. Few citizens then believed that the population would ever reach 120,000, and those who ,1,1 tn venture the onimon WCIC 1 1 1 1 1 civjufe.i that this limit would ever be exceeded, were regarded as dreamers. It was pointed out that the proposed capacity of the gravity works would suffice to give 60 gallons of potable water per head per day to 120,0) people, and as not more than four-fifths of the population would probably ever become consumers of the water, the supply was thntieht to be adequate for a great many years. i uese estimates, it must be remembered, were made twenty-three years ago, when the fund of information regarding water supplies and the growth of cities was not only very much smaller than at present, but was also vastly more difficult of access.

They were the outcome of many careful inquiries, and are, therefore, entitled to lenient criticism. Soon after the completion of the works, however, the city boundaries were moved far outwards on all sides. A large suburban population was thus added and the municipal territory was increased to over 11, fx) acres, with about 220 miles of streets. This was followed by a period of rapid development, so that Rochester has now an estimated population of 160,000, and an area of nearlv acres, with 295 miles of roadways, in which 253.72 miles of water distributing mains have been laid. Of this aggregate length 237.27 miles belong to the Hemlock Lake svstem and 18.46 miles to the Holiey system.

Connected with these mains are 2.4S2 hydrants, 2.480 stop-valves and 27,228 taps or service pipes. The capacity of the old conduit from Hemlock Lake to Rush Reservoir had beer, carefully gauged several times during the spring and summer of 1876, and found to be about 9,000,000 gallons per day, or 25 per cent, more than had originally been estimated. As a consequence of this large delivery, no restrictions in the use of water were enforced by the authorities up to 1SS5. In the meantime, the poor quality of the subsoil water and the drainage of many private wells by sewerage works in almost every portion of the town had caused such a demand for the Hemlock Lake water, that the distributing system was rapidlv extenrled, and soon almost everv building in streets where water mains were laid was provided with a water service pipe. The consumption thus became much larger than had been estimated to occur so soon after the works were fln-shed.

and as large quantities were lost by defective house fittings and lavish use. it finallv became necessary in 18X6 to enforce measures for preventing waste as it was evident that the limit of the conduit's capacity would otherwise soon be reached. PLANNING THE WORK. Absolute Necessity of a NeV Conduit Apparent Eight Years Ago. "With the continued fast growth of the city, however, these restrictive measures could not long be successful, and the limit was at last reached in when the population was about 125.000.

with ISO miles of mains and 21.0UO house connections from the Hemlock Lake system. In that year the 6torage in the reservoirs was nearly exhausted, and it to then definitely realized that the demands for the water were greater than the supply. Chief Kngineer Tubbs was therefore authorized to make surveys and examinations for procuring an additional supply, and in September of the same year, he presented his report in which, after considering various remedies for the existing conditions and a few Hources for the additional supply, he recommended the early construction of a new conduit with a capacity of 15,000,000 gallons per day, from Hemlock and Canadice The plans outlined In this report, however, did not appear to be generally accepted, and the matter was submitted on February 16, 1889, to the well-known engineers, Messrs. A. Fteley.

of New York, and J. T. Fanning, of Minneapolis, for thorough Investigation. Several new. surveys were made under their direction, and on 14, 189, they reported in favor of either Hemlock Lake or Conesus Lake as a proper source, and that the new conduit should have a capacity of 15.WVK gallons per day.

While their plans differed in detail from those of Mr. Tubbs, they agreed with the latter as to source and required quantity. No definite action upon these reports was taken by the municipal authorities, and the agitation of the subject was briskly continued by the promoters of other plans. Meanwhile It became evident that temporary relief would be needed, even if the construction of a new conduit to Hemlock Lake were authorized, since such an undertaking would require at least two years for its completion. A number of sources for temporary use were thoroughly examined in 18S9, and early in 10, but as they proved to be deiicient.

either in quantity or quality, the only resource left to prevent the depletion of the reservoirs was to maka the consuriptlon la the city equal to the jn turn embracing bids tor Dotn sien east-Iron pipe, and for two different routes designated A and Ii, for a portion of tiu distance. It should be mentioned that Route A lot the pipe conduit was located in the alluvia flats through which Honeoye Creek mean ders, and involved several difficult cross ings of the existing and former channels of said creek: while Route avoided some of these crossings bv following the course oi a certain lateral valley for some distance and then returning to the main valley by means of a masonry conduit and tunnel about 1.000 feet long. These two routes were each about 8,000 feet In length and formed a part of the entire length of 140,000 feet of conduit to be constructed in either case. The lowest bid for the masonry conduit and tunnel work was made by William H. Jones Sons, of Rochester.

N. its aggregate amount, as computed from the prices and estimated quantities, being that for the pipe conduit was made by the Moffett, Hodgkins Clarke of New York, their bids amounting to with 38-inch steel pipe, by Route and to $857,552.50, by Route for the same work with 36-inch cast-iron pipe as an alternative, the only bidders were Whit-more, Rauber Vicinus. of this city, and a syndicate of cast-iron pipe manufacturers, respectively, their joint bids amounting to $1,173,110.00 by Route and $1,177,035.00 by Route B. From the foregoing bids. It was seen that bv Route A the cost of the new conduit with 36-inch cast-iron pipe would be about 29.9 per cent, greater than with 38-inch riveted steel pipe, and about 37.3 per cent, greater by Route B.

Much surprise was caused by this result, as well as by the entire absence of competition on the part of manufacturers of east-iron pipe. It may also be remarked that the delivering capacities of the two kinds and sizes of pipe just mentioned were regarded as equal. Before the work was definitely awarded, inquiries were made of several manufacturers of large cast-iron pipe whether any lower bids for such material would probably be received in case this part of the work were re-advertised for letting. The replies were all to the effect that no reduction of prices could reasonably be expected which would make the cost of the work with cast-iron pipe nearly as low as that which would result by adopting the steel pipe; and as it was desirable that no further delays in commencing the work should occur, the decision was reached to construct the pipes of steel plates riveted together, and proierly protected from corrosion by an asphaltic coating. This decision was also intluenced in a large degree by many satisfactory reports as to the durability of such pipes in other localities, as well as by the good condition in which the old wrought-iron pipe conduit was found on its recent complete exposure for distances of several hundred feet at the crossings of the Lehigh Valley and Honeoye Valley Railroads.

The contract for the construction of the masonry conduit and its appurtenances, which is designated as Contract No. 1, was duly awarded to William It. Sons on January 6, 1893; and on January 12th, the award of Contract No. 2. for the construction of the 38-inch riveted steel pipe conduit, and ail the required appurtenances thereto, was made to the Hoffett.

Hodgkins Clarke Co. The proposals of these two firms were accompanied by thoroughly good bonds in the amount of $30,000 and respectively, conditioned that they would accept the contracts if the same were awarded to them, and hence there was every reason to anticpate that the documents would be executed promptly. The award was accepted by Jones fe Sons, but was declined by the other firm on the ground of an error in their figures. This of the Moffett. Hodgkins Clarke Co.

involved the forfeiture! of their said bond for $90,000. and to avoid this great loss, they quickly obtained from the 1 nited States Circuit Court a temporary injunction restaining the Executive Board of this-citv from declaring th'-m in default and proceeding with the collection of the money. Several attempts have since been made by the City Attorney to bring the case to trial, but in every Instance a wslponement was secured by the opposing counsel. The case la nf p-reiit imnortancp to the city and it is hoped that a settlement will soon be reached. It was evident, from the legal procedure referred to.

that great delay would resull before the said company could be com pelled bv the court either to abide by it bid. or submit to the forfeiture of its bond; and as such delay would be prejudicial the sanitary condition of the city. Contract No. 2 was awarded on January 20. lmci.

to the next lowest bidder, which was the firm of hi i more, li.HUbcr Vicinus. of thi city. The award was made on the basis of using slet-1 pip" with Route A. an was promptly accepted. The amount of tht proposal, as computed from the prices and estimated quantities, was $1,123,920.

BEGINNING THE WORK. Nearly Two Years Since the Contractors Began Active Operations. Early in March, 1893. both contractors for the new conduit commenced active operations on the lines of their respective work, the few preceding weeks been fully occupied with making the necessary preparations. Arrangements for the speedy manufacture of the steel pipes were then also concluded by Whltmore.

Rauber Vicinus, the work being divided equally between the East Jersey Pipe Works, of Paterson. N. and the Rochester Bridge Iron Works, of this city. The most Improved machinery and appliances for the purpose were placed In both of these establishments, and the manufacture of pipe began as soon as the steel plates arrived from the rolling mills. For the first-named works, the plates were made by the Carnegie Steel at Homestead.

and for the second, by the I'axton Rolling Mill at Harrisburg, Pa. At each mill every plate was carefully examined by a competent inspector, and frequent tests of the quality of the material were made, as provided in the specifications. A check on this part of the work was also constantly exercised by independent tests made in other lalxirato-ries, as well as by the city's inspector at the pipe shops, so that every possible care wl: taken to secure good and sound plates. The contracts provided that the steel should be of the class termed "soft," and be made by the open hearth process. Chemical analyses of each melt were required to show that the metal contained not more than 0.60 ier cent, of manganese and 0.6 per cent, each of phosphorous and sulphur; while the physical tests called for a tensile strength of between anil 65,000 lbs.

per square inch, with an elastic limit of not les3 than 30.0oo lbs. per square inch, and an elongation of 22.5 per cent, in a length of 8 inches; also for various proofs as to cold bending, punching, drifting and forging. Similar requirements were likewise made for the Iron and steel rivets. The thickness of every plate was determined by fine micrometer measurements around the edges, and all plates which were less than 95 per cent, of the required thickness at any point were rejected without appeal; furthermore, at least 90 per cent, of the accepted plates were of full thickness at all points. Every precaution was also taken to prevent the formation of rust upon the metal, and to this end it was constantly kept under cover, from the time of its manufacture at the rolling mill, until the completed pipe was thoroughly coated in the pipe shop with a durable asphaltic varnish.

In case of any accidental rusting, the plate was either rejected, or the rust was entirely re moved by manual labor. It was contemplated that the pipe conduit should be a continuous riveted tube throughout its whole length, without expansion joints or other mechanical devices for compensating changes in length caused by variations in temperature. By this continuity, such changes are prevented from manifesting themselves, and are replaced by internal stresses In the metal, which are far within its elastic limit for the range of temperature encountered in the work; but it becomes essential in this case that the strength of the net section of the plate at the circular seams shall be equal to the shearing resistance of the rivets therein. The joints were therefore carefully designed to meet this condition, and at the ends of the conduit, motion was prevented by bedding the pipe in large masses of masonry. Similar precautions were also observed in the design of all stop-valve and branches which were inserted in the pipe line.

All of these appurtenances were provided with flanges and bodices of considerably greater strength than the pipg, and were connected thereto by numerous heavy bolts. Great care was taken to make accurate maps and profiles of the route on a large scale. Upon these the pipe was carefully drawn, and every deviation from a straight line clearly indicated, after which they were handed to the manufacturer. As 11 was required that the pipe should be madf so as to bring all the straight seams in the upper quarter of the circumference, clost attention to the work on his part was Imperative, and a good check on the computations made by the engineering force war thus secured. In manufacturing the plates Into pipe, thf utmost care and precision in edging, punching, riveting and calking were used.

One half of the whole number was rolled lnti truly cylindrical "outside" sections oi cpurses, and the other half Into similar eijii. UoTn and all the others were six incUei Tu consisted of two atop-valves bolted and to a flanged casting riveted on the 1 quarter of the pipe, the valve nearest ll latter being kept open and regar.l.d reserve in case of injury to the other n. which Is ordinarily kept closed. An iron nh1 leading to the natural watercourse or age ditch, in the depression serves to the overflow, when the valve is opened some distance from the conduit At blow-off. both the main pipe ami the nair stop-valves are supported upon a bed of crete, so as to reduce the dauger of breakage in the connection from unequal 7 tlemeul.

1 Btl' Ten large and heavy 3G-Inch are placed in the line of the conduit Its t.itcrliiiilii,, i lOJiil iv. irwriujir, Ullil Mil quired through that portion of tlie vail. where the new conduit crosses nine ti, under Honeoye creek, three connections tween the two can be made so as to uiviil this somewhat hazardous part of the rout. Into two sections, either of which can l. out by means of these valves without gretl' Impairing the etliciency of either of if, Tu lines.

Provision has also been made for netting with the old conduit at two point the first at Kichmond Mills, where the oM pipe may be made to partially supply the new line, and the second near North lilooiuHelil where the relative elevations lirst heciuTl siK-h as to enable the new line to feed old one. There Is, howevvver, no necessity for making any of these conii'we. tlons. and facilities therefor have Iweu pro. vided only because It Is probable tlmt tln will become desirable at some future time The old and new conduits are also nected indirectly in the new gate Ihhism Kush reservoir, where a complexity of valvei Is found necessary for the purpose of avu'll.

ing grave annoyances in the future, whon another conduit from Hemlock hike Is lmli" The pipes and valves have here been arrant so as to allow any of the conduits comine to said reservoir to feed any of ihe plnei leaving the snme for Mt. Hope reservoir. similar provision has also been made at" point In Pinnacle avenue, about 4 feet norrd Klmwoisl avenue, where the present new pipe deflects abruptly to Mt. Ibipe reser. voir, the intention here being to treat tli deflected line as a branch of said reservoir while the conduit Itself Is continued to a new distributing basin, as previously lueiitiunrd Another large group of valves occurs la the new gate house at Mt.

Hope reservoir where the new conduit Is directly eoiineetrii with two new distributing main These latter pipes are nlso designed to taki water from the reservoir: bur ns it maw become desirable at times to obtain a koiiik. what higher pressure In the city than i afforded from said reservoir. It "lias hwn planned to equip each of these pipes ta a check-valve, which will close automatically when they are fed from Iiush reservoir, anS will re-open when the connecting valves art shut. A few words with respect to the ennnee. tlons of the new conduit with the I iri nr.

voirg may also be of interest. For the pre, ent. only temporary connections have Ih-b made by menus of a short line of lti-inrh pipe laid over the bank at each locality; Imt an the available pressure is large eiiuuih to deliver from ten to twelve million gallonn per day through both 1'nes, they are amply sufficient until next summer, when the permanent connections will be made. In the performance of this work, it will lie necen-sary to empty both reservoirs and keep the water out for some time. In order to mt through the banks, lay the various large inlet and outlet pipes, construct the masonry screening wells, and finally refill the excavations: and as it was expedient to have fnil assurance from long continued good service, that Ihe new conduit would not require shutting off for repairs during this period, the work was accordingly postponed.

Much more miirht be added to the foregoing by describing the extensive preliminary grading work which was done In several portions of the route before the pipes couM lie delivered, as well as the difficulties attending the completion of the crossings tinder Honeoye creek and the several lines of rai-wnv; but as snnee therefor Is not now nviil-able, and nothing has yet been said of th instruction of the masonry conduit nnd th" Intake pipe, these portions of the work will next be considered. DESCRIPTION OF THE WORK. The Aim Was to Secure "Stabmtytry Tunneling Through Rock. The reasons for adopting a diameter of feet and a grade of one In 4.UOO for the masonry conduit, from Hemlock lake to a point about 12,000 feet northerly, have already been Indicated, and we may therefore pass to the description of tlie work Itself. From numerous test lioriiiKs.

it found that the firmest subsoil, at the contemplated depth below the existing ordinary low water level, occurred over a cuinput'ii-tively small urea at the northeastern corner of the lake. This place was accordingly selected for the site of the gate house, un I the route of the conduit iu earth excavation was governed by similar considerations. It wns also aimed to secure the utmost stahiilt? for the work by reaching tunnel work In rock as quickly as was consistent with the shortest expedient length of line, since thert was little difference in the cost of the finished conduit, whether tunneling was done In earth or In rock. From the contract it wns found that when the depth of an open trench lu earth reached a limit of HO feet, the cost per lineal foot of the compli-tnl unlet re would be the same as If the work were done In tunnel, and hence this depth of SO feet became the dividing Hue between tut two classes of excavation. An examination of the profile of the routo shows that at the beginning of the conduit proper, which is aliout loo feet from th shore, the ground Is some four feet abort the ordinary low water level, and hence that the excavation would here be about ".2 bs't deep.

From this point northerly the ground gradually rises, until at a distance of feet, the depth of cutting becomes feet. Continuing along the route, the surface rises rapidlv to Shaft No. 1. which Is ti.3 feet deep, "and It does not again reach to within feet of the required bottom grade of tlie conduit for a distance of feet, in tlie remaining length of 4.1M feet to the end of the work, the depths of excavation are sil less than 30 feet, the average being feet. The contractors were aceordint'lr to tunnel the said distance of feet, and to make an open trench for tlie remainder of the wav, but as they preferred to continue the tunneling for some bevond the limits at the sime prices as fr the work in open excavation, uo ol'Jcctiuu was made.

To expcdlate the tunnel work, six snnrtt were sunk at nearly ciul-dlstaut Ia.V',: ably situated points nloug tne mine, affording, with tie two end points. Tou- wa 111 Cilliu nii'i ti.o t.oner for operating tne rH drills was compressed air. which was t1'' uruis was ruuii'nf mined through a long line of 4-inch Iron pipe, from a located station each of the shafts, and afirwr.l to the north portal. orK was menced on March 2. at Shaft No.

and soon afterward at the other five. this shaft was the shortest, its excsxai waws finished on April 1 Ith. when tun. lug wns first begun In both directions. to the north portal.

Work was first cer i. uiri oi shoft NO. 111 I- il. r. hist of the other shafts finished, and tunneling could tli.

from twelve points of attack. As tne terinl at one of the two portals was clay, while at the other soft ami friable rock curred. the excavation nt these delayed bv the contractors until the prin cipal rock" work was well under ay or portals was begun on August l.itli 24th. respectively, timbered shafts deep to grade being sunk at these l'1-From the south portal, the excavation feet northerly, when tne ejnin. passed wholly Into clay.

T1p work In oft rock and clay was nd both the roof ami sides of the. tut net Meanwhile, work on that portion of tl-e eo dult which Wos to be built open treat. Corresponding to the air-valves oft for emptying the pipe, iu Ca of cU or repair, was proviued in every denri-L' Uj One of these was a i I 1 fflllB II. Hope reservoirs. Of the former, five placed in such positions that If a nnraliU conduit aliout 2.5S miles in lenirth Is i 1 teen places at which excavation mmu i -secuted simultaneously.

The depths of ttiew shafts ranged from 54.7 to s.J eet. average being i7.2 feet, of which U.7 f'-t r. r.T i shale The tunnel work was proseeuic. otislv throughout the twenty-four hours each dav. except Sunday, at a maximum ratt of feet, and an average rate of per day.

at each of the twelve he .11 J.n AiiL'imt Mh. the headings met Shafts No. 2 and No. .1. and within i weeks the other headings met.

so that September 1. a continuous passage formed between Shafts No. 1 and N'- nr r. VT-. ti.ot Work nt ihe two enns Ultl II In clay southerly to tne gaie -v northerly for 210 feet, after which is n.th Into nnd through soft, shale rock for a 1.

n-' of 1T0 feet to hard and dense rock, rn the north portal fX'-i' also rfMK HH tMtf'iltli I t-l here reoulred heavy timbering ami In the firm rock, on the other hand the tr.r needed no support, an.1 but Id tie tr from the falling of material was e.1. Progress from the portals was. heref. re somewhat slower than from the shafts, tr. averasre rate being about fet per d.iT.

ret on October 1. ttlt entire distance of 7.3.V) feet between ttit ment was made with a private company whereby a large quantity of water was to be delivered bv them, at very reasonable rates, into tbe "distributing reservoir from a series of wells which they proposed sinking in the flats about one mile south of said reservoir. This source had been discovered bv the company a short time previously, and full control of the land for a considerable distance around their experimental wells had been secured by them. As no other arrangements for the use of this water could be made, the company accordingly caused six 10-inch wells to be drilled and equipped with suitable pumps, and a 16-inch cast-iron force main about 5,000 feet in length to be laid to the reservoir. The first delivery of water was made on June 12.

1893. at the rate of only about 500.000 gallons per day, and although every effort was made to increase the yield of the wells, no appreciable improvement occurred thereafter. This was a serious disappointment both to the company and to the public, since the volume thus delivered was by no means adequate to enable the Water Works Department to dispense entirely with the obnoxious system of balancing inflow and outflow from the reservoir. jAs this quantity, however, afforded a little relief, the use of the water was Continued until October, 1S94. THE SOURCE OF SUPPLY.

Comparison of the Lakes Available For the Purposes of the City. Early in October, 1S90, the writer was requested to investigate the various general plans for a permanent additional water supply which had previously been submitted by the other engineers above mentioned, and to make such further examinations for the same purpose as he might deem necessary. This work was soon undertaken, and on December 16 and 30. 1890. reports and estimates were presented for permanent works from Hemlock Lake, Conesus Lake and Lake Ontario on somewhat different plans from the others.

These documents also exhibited a comparison of all the various estimates which had been made up to this time for new works from the same three sources, and it was seen therefrom thet the choice lay between Conesus and Hemlock Lakes, both on the score of quality of water and cost of construction. A brief description of the three sources to which reference has been made in the foregoing may now be given. Hemlock Lake is a beautiful body of water about 6.6 miles long and 0.6 mile wide. It is situated in a hilly region about 29 miles south of Rochester, and lies about 38" feet above the general level of the city. Its area at ordinary low water is 1,828 acres, and its average depth in the middle is 65 feet.

Including the water surface, its drainage area is 27.554 acres, or 43.05 square miles, a large proportion of which is steep and covered with forest. The only settlement of considerable size on the watershed is the village of Springwater, which is located nearly three miles south of the head of the lake and has a population of about 600. On the rest of the area the resident population is proD-ably about 1,0. Canadice Lake lies about two miles east of Hemlock Lake, and is 2X) feet higher than the latter. It is about 3.13 miles long, 0.33 mile wide, and its average depth in the middle is 70 feet.

At ordinary low water it covers an area of 648 acres, and the total drainage area is S.8S3 acres, or 13.88 square miles. This territory is occupied wholly by farms and woods, with a population of not more than 400. The stream which forms the outlet of Canadice Lake must also be considered in this connection, as it unites with the outlet of Hemlock Lake at a point about one-third of a mile below the foot of the latter lake, and has a drainage area of 3,515 acres, or 5.49 square miles. The character of this territory is similar to the other watersheds, and Its population is about 200. It should also be noted that for nearly one mile below the foot of Hemlock Lake, the ground is nearly level along the course of the outlet, but rises with easy grades to the base of the hills which form the sides of the valley.

With such a eonliiruration of the surface, it is obvious that the flow from Canadice Lake and its outlet can readily be diverted into Hemlock Lake, either by means of a new open or closed channel, or by a low movable weir placed across the channel of the Hemlock outlet below its junction with the Canadice outlet. In this manner Hemlock Lake can be made the general reservoir for the yield of the three distinct watersheds already considered, whose aggregate area amounts to 39,952 acres, or 62.43 square miles. Conesus Lake lies parallel to, and a few miles west of. Hemlock Lake. It is about 7.8 miles long, 0.63 mile wide and has an average deptff of 45 feet in the middle.

At ordinary low water, its area is 3,184 acres, and its elevation is about feet above the general level of the city, or 78 feet lower than Hemlock Lake. Including the water surface, its drainage area is 39.980 acres, or 62.47 square miles, and is therefore practically the same in extent as that of Hemlock and Canadice Lakes and Canadice outlet combined. The character of the territory, however, differs from that of the latter by having a much larger proportion of open farming land, less abrupt hillsides and a greater average density of population. Several good-sized village, as well as a few smaller settlements, are on the watershed, and from the recent census statistics it may be estimated that the total number of persons residing thereon permanently is at least 3.500. This lake extends a few miles further north than Hemlock Lake, but from the preliminary surveys it is found that the length of a practicable route for a conduit to Rush Reservoir is only 2.6 miles shorter than one from Hemlock Iake.

Lake Ontario is about seven miles directly north of Rochester, tut its average water surface is 2tk feet below the general level of the city, and 290 feet below that of Mt. Hope Reservoir. To avoid the influence of the roily and sewage-polluted water of the Genesee River, the intake of a conduit for supplying the city must necessarily be located as far distant from the river as possible. As the prevalent winds are from the west and northwest, and from the peculiar configuration of the lake shore, it Is obvious that no suitable location for an intake can be found at any reasonable distance easterly from the mouth of the river, and hence such a location must be sought to fee west thereof. For some miles in this direction, however, a series of swampy bays or ponds, connecting with the lake, are encountered, whose waters are entirely unfit for domestic use, both by reason of excessive aquatic vegetation, and from the fact that they are contaminated by the drainage A.