TRANSACTIONS OF THE SECTIONS. 25 



been projected in a curve. This curve was shown to the meeting. It is evidently a 

 continuous curve, convex towards the valley, and not a zigzag motion, such as must 

 have resulted from distinct rents parallel to the length of the glacier. The length of 

 the line, originally straight, whose deformation was observed, was 90 feet, and the 

 ordinates of the curve were determined by accurate measurements at forty-five 

 stations two feet apart. The experiments on the continuous flexure of the transverse 

 line were extended over a longer period, at points 30, 60, 90, 120, and 180 feet from 

 the theodolite, with similar results. 



The author concludes, — 1st, that the sliding of the mass of the glacier over itself by 

 insensible gradations cannot be denied, and that it is sufficient to account for the 

 observed excess of progress of the centre above the sides of the glacier ; 2nd, that this 

 diflferential motion takes place in the direction in which the veined structure exists, 

 and that it is impossible not to consider the one pheenomenon as dependent on the 

 other. 



Experimental Inquiries i?i(o the Falling -off from perfect Elasticity in Solid 

 Bodies. By Eaton Hodgkinson, F.B.S. 



At the Cork Meeting of the British Association, Mr. Hodgkinson laid before the Sec- 

 tion the results of some experiments, the object of which was to show that no rigid 

 body is perfectly elastic ; the slightest change of form in a body producing a perma- 

 nent alteration in it, however small. He endeavoured to show, too, that in experi- 

 ments on the deflexion of rectangular bars of cast iron, and some other materials, the 

 defect of elasticity, denominated the set, varied as the square of the weight laid on, 

 nearly. It might, he stated, be inferred, too, that the set varies as the square of the 

 deflexion, since the deflexion is as the weight nearly, though it varies in a ratio 

 somewhat higher than that. 



In rectangular bars bent so as to strain them in a small degree only, the particles 

 are equally extended and compressed on the opposite sides of the bar ; but in bars 

 whose section is of the form A-pB, the deflexion arising from the flexure of the plate 



C 

 A B and the extension or compression of the part C, varies in a higher degree than as 

 the square of the weight, and in these the set varies nearly as the square of the de- 

 flexion. 



Mr. Hodgkinson stated that, soon after the Cork Meeting, he had received, from 

 a very intelligent writer, a letter on the subject of the communication here described. 

 In this letter considerable doubt as to the correctness of his conclusions was expressed, 

 and it was suggested that the facts might probably be accounted for by attributing 

 them to friction between the ends of the beam and the supports on which it rested, 

 a matter which had been investigated by the Rev. Professor Moseley, in his able work 

 on the Mechanical Principles of Engineering and Architecture (Art. 389). 



Mr. Hodgkinson felt convinced that the causes mentioned in the letter were not 

 the right ones, but thought it incumbent on him to obviate, as far as possible, all 

 objections, and to show that friction was not the cause of the results observed. 



In his former experiments the weight of the bar was neglected, as it was very small 

 compared with the weight laid upon it ; and the deflexions and sets were measured 

 from that position which the middle of the bar had taken in consequence of its own 

 weight. The friction upon the ends of the bar, from the supports on which they 

 rested, had likewise been neglected ; and the quantities of the sets, usually very small, 

 had been measured by an instrument (a long wedge graduated along the side) ; and 

 although this was good comparatively with some previously used, it did not admit of 

 all the accuracy which was required. 



He had, therefore, an apparatus constructed to remove these defects. In this ap- 

 paratus, the bar or body bent is laid upon its edge or smallest side, and the force to 

 bend it acts horizontally. The ends of the bar are supported horizontally and ver- 

 tically by friction rollers, and the deflexions and sets are measured from the centre of 

 the " straight edge," in which screws resting on the ends of the bars opposite the 

 rollers are inserted. The sets, and the smaller deflexions, are measured by a micro- 

 meter screw, in the centre of the straight edge, capable of measuring distances as 

 small as xooooth of an inch. In this apparatus the flexure of the bar, being horizontal. 



