Abstract
LONDON. Physical Society, January 26.—Prof. Lodge, F.R.S., President, in the chair.—A paper by Prof. Ayrton and Mr. Mather, on some developments in the use of Price's guard wire in insulator tests, was read by Prof. Ayrton. For insulation tests made by the direct deflection method the guard wire properly applied affords complete protection against surface leakage when the ends of the cable tested are near the galvanometer, so that it is possible to have the wire connecting the conductor of the cable with the galvanometer terminal “air insulated.” A difficulty, however, arises when the ends of the cable are at a considerable distance from the testing instrument; this may render air insulation impossible. The authors have overcome this difficulty by applying a guard wire along the entire length of the lead. This is done by using a concentric wire to connect the cable and galvanometer, the inner of the concentric being used as the lead and the outer as the guard wire. The principle can also he applied to determine whether a defective piece of cable is bad throughout or bad owing to one or more isolated faults. In this case the cable is placed in two water tanks, one of which is earthed, and the other fairly well insulated. By a suitable arrangement of the guard wire it is then easy to determine the resistance of the wire in the earthed tank, so that by altering the length of this wire the character of the insulation can be determined throughout the whole length of the cable. In referring to some of the earliest experiments with the guard wire made by Mr. Appleyard in 1895, Prof. Ayrton pointed out that the principle had not been applied completely, and that at one point there was a chance of leakage. Mr. Campbell said that the necessity of having a concentric could be obviated by simply hanging the lead from the guard wire by short lengths of material of fair insulation. Mr. Appleyard said that he quite agreed with Prof. Ayrton that the guard wire ought in general to be applied at both ends of all leads, provided that both ends could be got at. The reason it was used at one end only in the experiments on dielectrics made in 1895 was that the far end of the lead was carried into the condenser box, which was submerged in water in the temperature tank. Special precautions were taken to ensure good insulation of the submerged end of the lead, and tests showed that the leakage there was nil. As the end of the wire could not be got at, no guard wire could be applied. Mr. Appleyard congratulated the authors upon the use of a concentric cable for a lead, and pointed out that such a lead was sufficient for all the routine tests on core; the inner and outer conductors could be used for the purpose of taking the “opper” resistance. Mr. Price expressed his interest in the developments of his principle which had been made by the authors.—Mr. Appleyard then read a paper on a fault-test for braided and other cable-core. This method enables the fault to be found without the removal of braiding or tape. The core is wound on two insulated drums or tanks, the intermediate piece of cable being about ten feet long. One end of the core is left free, the other is connected to earth through a galvanometer and a battery. A guard wire is connected from some point between the galvanometer and the battery to some point of the braiding on the wire between the drums. A wet cloth, connected to an earth wire, is laid on one or other of the drums, over the braiding. The galvanometer deflection is noted. The earth-wire is then changed over to the second drum, and the corresponding deflection is observed. A comparison of these deflections at once indicates upon which drum the fault lies. With the galvanometer still deflected, the core may be run through a suitable contact brush or sponge attached to the guard wire. The instant the fault passes under the guard wire contact, the deflection falls and the fault is located. The paper gives the theory of the method, and indicates how to apply it (1) to localising “distributed” faults; (2) to several faults in a single cable; and (3) to the case of a single fault. One advantage of the method is that at the critical moment, when the fault passes under the guard wire, the galvanometer is short circuited through the fault, and thus completely protected.— A paper on reflection and transmission of electric waves along wires, by Dr. E. Barton and Mr. L. Lownds, was read by Dr. Barton. The waves used were produced by means of an induction coil and an oscillator, and travelled along wires.15 cm. diameter, 8 cms. apart, and 166 metres long. The ends of the wires were connected by graphite markings on ground glass, so that any wave trains which reached the ends were at once absorbed. Three circular parallel-plate condensers were used, of 15, 9 and 5 cms. radius respectively. The plates were in all cases separated by air, and were placed 1 cm. apart. The needle of the electrometer connecting the wires waVuncharged, so that it was always attracted by the charged plates. The positions of the condenser and electrometer could be varied so as to study either the reflected or the transmitted waves. The electrometer produced a negligible disturbance, as it reflected only 0.04 per cent, of the energy incident upon it. The authors have attacked.the problem mathematically, using the relations of Heaviside, and have obtained expressions for the reflected and transmitted systems. These expressions consist of two terms, one of which is comparatively unimportant. From the other term certain values have been calculated. A superior limit has then been given to the other term, and the values already obtained have been subjected to a correction on this account. By a suitable arrangement of the condenser and electrometer these calculated values have been experimentally determined, and are in close agreement with the theoretical numbers, falling in many cases between the results derived from the approximate and the corrected theories. The authors have also investigated the stationary wave system produced by interference when the electrometer is placed close to the condenser, and between the condenser and the oscillator. The chairman saic? that the experiments afforded a satisfactory verification of Heaviside's theory.—A paper on the frequency of transverse vibrations of a stretched india-rubber cord, by Mr. T, J. Baker, was taken as read. In this paper Mr, Baker has investigated the frequency of the note given out by an india-rubber cord of square section when subjected to different tensions. The relation between length and tension is linear over a considerable range. The curve connecting length with frequency shows that while the cord was doubling its length the pitch was rising rapidly, but that further extension was practically without effect. Since the relation between length and tension is linear, while the sectional area is decreasing, it follows that the value of Young's modulus must be changing. The author has shown that the value of Young's modulus is proportional to the square of the stretched length of the cord. Using this fact, the frequency of the note given_out by a stretched india-rubber cord is shown to be proportional to a quantity which varies very slightly with increase in length of the cord, and hence the variation in elasticity is given as the cause of the constancy of the note.mdash;Mr. Appleyard exhibited some mirrors produced inside incandescent lamps by the application of voltages much above those for which the lamps were designed, and the consequent deflagration of the filaments.—The meeting, then adjourned until February 9.
Article PDF
Rights and permissions
About this article
Cite this article
Societies and Academies . Nature 61, 334–336 (1900). https://doi.org/10.1038/061334a0
Issue Date:
DOI: https://doi.org/10.1038/061334a0