Recent Proceedings of Societies. 



Engineers' club, Philadelphia. 

 April 17. — Mr. H. W. Sanborn made some remarks 

 on recent stream-gauging for the future water-supply 

 of Philadelphia. The streams gauged were the Per- 

 kiomen Creek and tributaries in Montgomery county, 

 the Neshaminy and tributaries and the Tohickon in 

 Bucks county. The original intention was to gauge 

 the minimum flow only, and for that purpose weirs 

 were constructed on eight different streams. They 

 were very substantially built, as they had to withstand 

 the run of ice in the spring of the year. Heavy 

 bed-logs were placed at the level of the bed of the 

 stream, and the superstructure built on that. They 

 were made water-tight either by sheeting placed be- 

 low the bed-log, to rock-bottom, or a cement mortar 

 wall. The crests of the weirs were generally about 

 two feet above the beds of the streams, and were 

 made of two inch oak-plank. Gauge-boards were 

 placed about five feet above and below the weirs, 

 and connected, by levels, with the same. The one 

 above indicated the depth of the water on the crest. 

 The one below was used only in case the weir was 

 submerged by high water. The weirs varied in 

 length from fifteen to seventy feet, according to the 

 width of the stream. Stream-gauge stations were 

 established near the weirs. Headings were taken 

 there at the same time that they were at the weirs. 

 When a sufficient number of readings, at various 

 heights, were made, a ' curve of flow ' was plotted 

 by a comparison of the two. Then, when the crests 

 of the weirs were removed for the winter, the flow 

 was found by referring the stream-gauge readings 

 to the ' curve of flow.' The great fluctuation in the 

 flow of the streams, caused by the great number of 

 mills on them, necessitated a great many observa- 

 tions at the weirs to get a correct gauging. This dif- 

 ficulty was overcome by the use of automatic gauges. 

 They were run by clock-work, and drew a line on a 

 roll of paper, corresponding to the rise and fall of the 

 stream. The minimum flows were found to be so 

 small that the larger flows had to be determined. 

 These had to be found by other methods, for the 

 weirs would only carry, at the most, two feet in 

 depth, while the water in the streams sometimes 

 raised as high as sixteen feet. The measurements of 

 the large flows were made mostly by the use of elec- 

 tric-current meters. The measurements had to be 

 made from bridges, and, where none existed, in 

 proper places, small suspension-bridges were put up. 

 One was built over the Perkiomen, at Frederick, of 

 120-feet span, and one over the Neshaminy, at Kush 

 Valley, of 133-feet span. By means of the meter, 

 the velocity of water was taken at a great number 

 of places in a line across the stream, and a close 

 estimate of the velocity of the whole cross-section 

 determined. Stream-gauges were placed near the 

 meter-stations, to be read when measurements were 

 made, answering the same purpose as those con- 

 nected with the weirs. In some cases, large flows 

 were measured by getting the velocity of the stream, 

 by means of pole-floats. When used, care was taken 

 to have the length of them as near the depth of the 

 water as possible, and they were run at as many 

 stations across the stream as was necessitated by the 

 changes in the even flow of the stream. The rise 

 and fall of the water during freshets was so sudden, 

 and the stations, eleven in number, were so scat- 



tered, — the water-sheds covering five hundred 

 square miles, — that it was impossible to get to, and 

 make measurements of, more than one or two streams 

 during a freshet. Then, many times, the freshets 

 would come in the night, and nothing could be done 

 but the taking of continuous readings of the stream- 

 gauges. To overcome these difficulties with our 

 small force, and get at least fair measurements of 

 all the streams at the high point of a freshet, ' max- 

 imum stream-gauges ' were set up on most of the 

 streams. A place was chosen where the bed of the 

 stream was uniform in width and slope, and two 

 similar gauges set up. They were usually from two 

 hundred to five hundred feet apart. They were 

 made in the form of a box from eight to twelve feet 

 long, and six inches square inside. One side opened 

 as a door. They were placed on end and shielded 

 and supported by heavy timbers, embedded in the 

 soil or bolted to the rock bottom. "Vertically through' 

 the centre of the box ran a brass rod, which was 

 graduated. A metallic float ran on the rod in such a 

 manner that it would rise with the water, but would 

 remain fixed on the rod, at the highest point the 

 water reached, after it had fallen. The two gauges 

 were connected by levels, and from the gauge-read- 

 ings the slope of the water was determined. From 

 this the velocity of the stream was found by the 

 Kutter formula. The daily flows of all the streams 

 have been tabulated, from the commencement of the 

 gauging in July, 1883, to January 1, 1886, and the 

 field is still being continued. The daily flows have 

 also been shown graphically on sheets, with the rain- 

 fall on the watershed and the temperature annexed. 

 The connection between the three is well shown. 

 Eain-gauge stations were established over all the 

 watersheds ; and the data obtained from them, com- 

 bined with that from previously existing gauges, 

 which was kindly furnished us by the observers, 

 have also been plotted graphically, showing plainly 

 the variations of the rainfall over large areas. Three 

 automatic rain-gauges were used to show the inten- 

 sity of the storms. Mr. E. V. d'Invilliers spoke 



upon the geological position, characteristic features,, 

 and method of mining the ore at the Cornwall iron- 

 mines, Lebanon county, Penn. The ore-deposit 

 occurs in three hills, five miles south of Lebanon. 

 The extreme length of this magnetic ore-deposit is 

 4,400 feet in a general east and west line, and its 

 area is about 63 acres. The ore is surrounded on 

 three sides by a steeply sloping wall of dolerite (trap) 

 rock 100 feet ± thick, the mesozoic sandstone 

 abutting against the south-west dipping-ore on the 

 south side of the deposit. The ore was referred to 

 the lime-shale layers between the Siluro-Cambrian 

 limestone and the Hudson River slates, is magnetic, 

 practically free from phosphorus, but contains con- 

 siderable sulphur and some copper ; and, except in 

 the soft surface ore, all requires roasting before it is 

 worked in the furnace. There are three commercial 

 grades of ore ; but the bulk of the output is the No. 

 3 1 select ore,' mostly lump, with about 48 per cent of 

 iron and 2.5 to 3 per cent sulphur. Mining at pres- 

 ent is carried on entirely above water-level, though 

 the records of several bore-holes have established 

 the great depth of this deposit beneath the water- 

 plane ; one bore-hole being down 325' below the sur- 

 face, without any trap or other foot vein being 

 struck. The ore is mined in successive terraces and 

 stopes, as in huge open quarries, and, by means of 

 six compressed-air drills, large quantities can be 



