The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

144 uous deflection, and it will be seen that the amount of deflection will depend on the time the two brushes are simultane- ously on the metallic part of the wheels. This in turn will be proportional to the deflection of the shaft between the wheels, and hence to the power trans- mitted." The amperemeter could be graduated, if desired, to read directly in horsepower for a given diameter of shaft, number of revolutions, and dis- tance between the wheels. The author has not been able to get any data as to the practical performance of this ap- paratus, but the idea to him looked at- tractive. Invention of Archibald Denny. The type of torsionmeter with which the author has had a good deal of ex- perience, as already stated, was that in- vented by Mr. Archibald Denny, of Messrs. Denny and Brothers, Dumbar- 'ton, and Mr. Charles Johnson, a. mem- ber of their staff. The apparatus con- sisted of two wheels of non-magnetic material, usually gun-metal, fixed to the shaft at each end of the test length, each wheel being provided with a ra- dial slot in which a chisel-pointed per- manent magnet was fitted, the tip of magnet projecting about % in, beyond the rim of wheel. Fitted in a conven- ient position, adjacent to the rim of each wheel, was an inductor consisting of a soft iron core, wound on which were one or more coils of fine, insulated wire. The face of the inductor was curved to the same radius as the point of magnet, and was fixed at a distance of 1-16 in. from same. The magnet in passing these coils once per revolution generated an electric current in them by electro-mag- netic induction. The coils in each in- ductor were connected by means of a multicore cable to a recorder box, in which suitable contact keys were pro- vided, enabling the operator to make contact with any individual coil on in- ductor. The currents from these coils were led by means of the keys already mentioned to a differential telephone receiver. When the currents. did not reach the telephone at the same instant a _ ticking sound was heard. The keys were then moved till no sound was heard, which indicated that the currents reached the telephone at the same instant of time, or were "in phase". The sum of the readings on the keys connected to the two indi- cators gave the total deflection in inches, and from the simple empirical formula derived from that already given the horsepower could be obtained: Let-- N=Number of revolutions per minute, L=Length of shaft. THe Marine Review 153==A constant obtained from the modulus of rigidity. D=Diameter of shaft. R=Torsionmeter reading. C=Inductor constant. 1.53 R D*N Then H. P. = CL Proceeding to treat of the subject of power measurement by torsionmeters generally on board ship, the author stated that when using any form of torsionmeter it was advisable to check the zero reading when actually under way. He had known of cases where the alignment of the zero was perfect- ly correct when the shaft was at rest, but, through the turning gear having been left on, and there being a consid- erable amount of friction in the stern tube, the shaft, although at rest, was left in a state of twist. This was shown very clearly, when the turning gear was put in operation, by the shaft revolving in the stern tube with a series of jerks. The testing of zero readings could be easily carried out when under way in a twin or triple screw vessel by shutting off the steam from one shaft at a time, and allowing the propeller to be ro- tated by the action of the water. The shaft would then only be transmitting the power necessary to drive the tur- bine when empty, which was practically negligible. The readings could then be taken and compared with those taken when the shaft was at rest. These readings should preferably 'be taken both before and after the actual trials. Advantage Over: Steam Indicators. One advantage of torsionmeters over steam indicators was that they could be left in position permanently, and read- ings taken at any time, on practically a moment's notice. This was particularly true of the electrical type by which the actual measurement was taken at a station remote from the shaft tunnel. From the ratio of the power developed to the knots logged the observer was able to compare the combined efficiency of the combination, and also judge as to the effect of the ship's bottom when clean or foul. One very interesting point was brought out on the trials of a steamer some time ago. This vessel was fitted with four shafts, all turning inwards. The propellers on. the inner shafts were in the usual position for quadruple screws, the propellers on the wing shafts being a considerable dis- tance forward of the inner screws. When steaming at full power, with throttle valves full open, the inner pair of shafts was found to be developing the greatest power. This, on the face of it, was hardly what might have been expected, considering that the outer pro- June, 1909 es pellers were working in what was us- ually spoken of as solid water, while the inner screws were encountering water al- _ ready broken up by the propellers in front. The center shafts rotating as they did towards each other, and being closer together, seemed to act in such a way as to develop most power. In more re- cent ships, .the author believed, the shafts had all been arranged to turn outwards, and a very equal distribution of power among the whole four shafts had been: secured. In connection with the wing shafts an interesting point had arisen, and beer studied, which was con- cerned with the action of the propeller blades when passing, in the course of their revolutions, close to the ship's side. It had been thought that the varying dis- tance of the three blades throughout one revolution--or through one-third of a revolution--would cause an inconstant rate of work on account of the water being thrown against the ship's side more at one point of the revolution than at another. One firm of Clyde builders had carried out tests. with this in view, and taken readings of the twist at points 30 degrees apart, throughout one-third of a revolution, and they found no ap- preciable difference between them. This seemed conclusively to prove that the action of propeller blades could be ne- glected, and that the twist could be properly read at any point throughout the revolution. After . brief reference to the uses of torsionmeters in connection with en- gines other than those for marine pro- pulsion, the author, in closing, expressed the hope that his remarks had been suffi- ciently clear to show the great ac- curacy and adaptability of this method of measuring power. The ease with which it could be carried out, and the simplicity of the calculations required, all combined to render the torsionmeter an ideal means for the measurement of power, and there seemed no doubt of its extensive use in the future. In his annual report to the New York state legislature F. C. Stevens, state superintendent of public works, recom- mends such improvement of the outlets of both Cayuga and Seneca lakes as shall make these lakes a part of the new barge canal system when it shal! have been completed. From such data as he had at his disposal Supt. Stevens estimated that the deepening of the Sen- eca river, the Cayuga lake outlet, from the point where the barge canal enters the Clyde river, to deep water in Cayuga lake, can be accomplished for about $1,750,000 and that the deepening of the Seneca river from where it joins the waters of Cayuga lake to Geneva, will cost about $4,000,000.