194 



SCIENCE. 



[Vol. XIII. No. 319. 



GRAY'S TORSIONAL TESTING-MACHINE. 



The accompanying figure illustrates an apparatus recently de- 

 signed by Professor Thomas Gray of the Rose Polytechnic Insti- 

 tute, Terre Haute, Ind., for the purpose of testing the torsional 

 rigidity of different kinds of materials. 



The figure has been prepared from a photograph of a rough and 



inexpensive form of the machine, which was somewhat hurriedly made 

 in the workshops of the institute by students, for use in the engi- 

 neering laboratory course of the current year. 



The apparatus, as here shown, consists of a wooden trestle, on 

 the top bar, D, of which there is mounted a cross-beam A, about 

 four feet in length, which rests, through knife-edges at its centre, on 

 a support which can be clamped at any point of the bar D. The 

 ends of this beam are cut to circles having the knife-edges as centre ; 

 and to one end a thin steel trap, C, is fixed, the lower end of 

 which is attached to a cross-beam, F, of the same length as A. 

 The beam F\^ clamped to one end of the specimen, G, which is 

 being tested by means of strong clamps, which take different forms, 

 and are made of different materials, according to the form and 

 nature of the specimen. The other end is held in a similar clamp 

 at H, and this clamp is firmly fixed to the trestle. 



The end of the specimen to which the beam F is attached is kept 

 in position by means of an attachment similar to the tail-stock of a 

 lathe, the clamping-screw for which is shown at /. 



This centre-bearing also prevents any cross-bending force being 

 applied to the specimen by the weight of F. The torque, or twist- 

 ing-couple, is applied to the specimen by hanging weights, K, K, 

 on the free ends of the beams A and F. These weights should be 

 of equal amount, as they then produce a pure twisting-couple with- 

 out applying any force to the centre-bearing. 



The amount of distortion produced by any torque applied to the 

 specimen is measured by means of two indices E, E, which are 

 clamped to the specimen at a measured distance apart. The outer 

 ends of these indices carry a graduated arc, on which the angular 

 displacement can be read by means of a fixed mark or vernier. 

 For specimens of such large diameter that the limit of elasticity is 

 exceeded before a sufficiently large deflection can be given to the 

 indices E, E to render this method sensitive enough, the deflection 

 is indicated by a multiplying index, B. An important feature of 

 this apparatus is the elimination of any uncertainty as to effect of 



the clamps by measuring the relative twist at two sections a short 

 distance from the ends. 



This same method was adopted some years ago by Professor 

 Gray, in a series of experiments on the elastic constants of rocks, 

 but the apparatus was not then made in a permanent form. A 

 considerable extension of the experiments is now contemplated in 

 connection with investigations in seismology, under the direction of 

 Professor Mendenhall, in which it is intended to determine the 

 elastic constants of a number of rocks, for the purpose of ascer- 

 taining the theoretical velocity of a seismic wave. 



In the more complete design of the testing-machine above de- 

 scribed, both ends of the beam A are connected by straps or links 

 to the beam F. The tail-stock centre-bearing is then omitted, and 

 cross-bending stresses are avoided by mounting the clamp H oa 

 gimbals, which allow freedom to transverse motion. A graduated 

 disk is then substituted for one of the indices E, E ; and the other 

 index is carried on a bar which extends from the clamp, in a direc- 

 tion parallel to the axis of the specimen, up to the front of the 

 graduated disk. The relative distortion is thus read off direct 

 when that method is sufficiently sensitive, or by means of a second 

 index attached to the disk when higher sensibility is desirable. For 

 some purposes the gimbals are mounted on a worm-wheel, which 

 turns round an axis parallel to the direction of the specimen, which 

 thus allows an unlimited amount of twist to be given to the speci- 

 men. This becomes necessary when torsional strength is the ob- 

 ject of investigation. 



With the apparatus here illustrated, specimens of any length up 

 to three feet can be included between the clamps ; while specimens 

 of any length can be tested in sections of three feet or less, the 

 ends being simply allowed to project beyond the clamps, and the 

 tail-stock bearing modified to a V instead of a centre-bearing. As. 

 regards the power of this machine, it is capable of testing a three- 

 inch steel shaft up to its limit of elasticity. 



THE MACR^ON SECONDARY BATTERY. 



There is no field in which experiment is being more actively 

 prosecuted than in that of the storage of electrical energy. From 

 the experience which has been gained in the last five or six years,, 

 the failings of secondary batteries have become pretty well under- 



stood, and many inventors are trying to remedy them. The two 

 types of battery which have been at all generally used are the 

 Faure and the Plante. In the former a support-plate is provided, 

 and some salt of lead is mechanically applied to it, which forms 

 the active material. In the latter the active material is obtained 

 from the support-plate by reversing the current passing between, 

 two lead plates in dilute sulphuric acid. The Faure cells take but 

 a very short time to manufacture : the Plante type takes several 



