262 



NATURE 



[July i6, 1891 



tance {x) from the surface. The formula thus arrived at was 



p 

 y = o'64— , whilst for a finite beam centrally loaded the 



author's experiments gave y = 0726 — The experiments 



were made on glass beams mounted in a steel straining frame, 

 and placed between the crossed Nicols of a polariscope. Steel 

 rollers 2 mm. in diameter served as supports, and the central 

 load was applied by a screw acting on a roller of similar diameter. 

 Deflections of the beam were measured by a micrometer screw 

 at a point opposite the central load, and traversing screws enabled 

 the whole frame to be moved so as to bring any portion of the 

 beam in the field of view. Circularly polarized light was some- 

 times used, and a micrometer eye-piece served to measure the 

 distances between interference fringes produced by loading. By 

 carefully chosen experiments the author had shown that if a 

 beam of glass be laid on a flat surface and loaded across its 

 upper surface, the shear at any point on the normal at the point 

 of contact of the load is inversely proportional to the distance 

 from the point of contact. In the first experiment the crossed 

 Nicols were set at 45° to the axis of the loaded bar ; a quarter- 

 wave plate was then placed between the bar and the analyzer, and 

 the position of the black spot at the point where the effect of the 

 shear on the polarized light was equal and opposite to that pro- 

 duced by the quarter-wave plate was noted. A second quarter- 

 wave plate was then superposed on the first; the black spot 

 moved upwards to a point where the shear was double that at 

 the first position. This position having been determined, 

 one quarter- wave plate was removed, and the load diminished 

 until the original spot moved up to the second position, and 

 the processes repeated. By this means a series of positions at 

 which the shears were in the proportions i, 2, 4, 8, &c., were 

 determined. Plotting the results showed the curve connecting 

 the shear and the distance from the point of contact to be hyper- 

 bolic. Other experiments showed that the shear at any point 

 was proportional to the load. By maintaining a constant load 

 and measuring the distances between the interference fringes 

 below the point of contact the hyperbolic law was confirmed. 

 The effect of bending a beam is, according to hypothesis, to put 

 the upper portion in longitudinal compression, and the shear 

 (vertical stretch) varies as the distance from the centre of the 

 beam ; the shear due to surface loading is a vertical squeeze, 

 and, as shown above, varies hyperbolically. When, there- 

 fore, the beam is subjected to both actions, the straight line 

 representing the bending strain may intersect the hyper- 

 bola representing the shear due to surface loading in two 

 points, and since, at the corresponding points in the central 

 section, the shears are equal and opposite, the elements are 

 only subjected to voluminal compression, and will exert no 

 bi-refringent action. Hence, when viewed through crossed 

 Nicols, black spots will be seen on a white field. Keeping the 

 lead constant and diminishing the span should cause the spots 

 to approach each other, and when the line is tangential to the 

 hyperbola, the spots coincide. These deductions were con- 

 firmed by experiment, and it was found that for a span of less 

 than four depths, no point of zero shear exists on the central 

 section. The strains in beams subjected to surface loading were 

 thus shown to be of a character different from those usually 

 assumed, the neutral axis instead of coinciding with the axis of 

 the beam, being lifted up in the centre, and its shape depending 

 on the load and span. Other ingenious and interesting experi- 

 ments on beams were described, in some of which the lines of 

 principal stress were mapped out. Remarkable results were 

 obtained, showing that although the tension lines given by 

 Rankine and Airy are nearly correct, the curves of compression 

 may be very different, and have very curious shapes. Prof. 

 Perry thought the local loading effect would not be so important 

 in long beams, and inquired whether in ordinary test pieces local 

 loading would affect the breaking strength. He also asked 

 what effect the fact of the load making contact over a surface 

 instead of along a line would have on the results, and in reply 

 Prof. Carus-Wilson said the effect was to raise the asymptote of 

 the hyperbola representing the surface loading stress above the 

 surface of the beam.— On pocket electrometers, by C. V. Boys, 

 F. R.S. This communication described modifications of electro- 

 meters adapted for portability. As quartz fibres increase the 

 delicacy and diminish the disturbing influences affecting instru- 

 ments, much smaller controlling forces can be employed than 

 when silk is used for suspensions. He had, he said, pointed 

 out some time ago the great advantages arising from making 

 NO. I 133, VOL.44] 



galvanometers small. Applying similar reasoning to electro- 

 meters, he remarked that making an instrument one-tenth the 

 size of an existing one reduced the moment of inertia of the 

 needle to ^l^, whilst the deflecting couple for given potentials 



would only be yV of its former value. The small instrument 

 would for the same periodic time be 10,003 times more sensitive 

 than the large one, provided the disturbing influence could be 

 reduced in the same proportion. This, however, was not 

 ordinarily possible, for any method of making contact with the 

 needle, such as by a fine wire dipping into acid or mercury, 

 prevented very small controlling forces being used. Still, by 

 suitable devices a large proportion of the full advantage could 

 be obtained ; a freely suspended needle without liquid con- 

 tacts was essential to success. The first instrument described 

 was one in which the needle was cylindrical, contiguous quarters 

 being insulated and connected to the opposite ends of a minute 

 dry pile placed within the needle ; opposite quarters were thus 

 at the same potential, and at a different potential to the other 

 pair of quarter cylinders. This was suspended within a glass 

 tube silvered on the inside and divided into four parts by fine 

 longitudinal lines. In such an instrument the needle and 

 quadrants are reciprocal, and the deflection depends on 

 the product of the difference of potential between the 

 quadrants and that between the parts of the needle. Owing 

 to the dry pile not being constant, the instrument was found 

 untrustworthy, but when working at its best a Grove cell would 

 give 30 or 40 millimetres deflection. The next step was to 

 make a cross-shaped needle of zinc and platinum, and rely on 

 contact electricity to keep the parts of the needle at different 

 potentials. This bold experiment proved remarkably successful 

 for the instrument was very sensitive. A disk-shaped needle 

 with quadrants, alternately zinc and platinum, was then em- 

 ployed, and by this a small fraction of a volt could be measured. 

 The weight of the disk was only ^^o of a gramme, and the in- 

 strument could be turned upside down or carried about in the 

 pocket with impunity. Another small instrument with the 

 stationary quadrants of zinc and copper was exhibited, and by 

 rotating them through an angle of 90°, so as to bring 

 them in a different position relative to the parts of the 

 needle, a deflection of several degrees of arc was produced. 

 In the course of his remarks Mr. Boys made several sugges- 

 tions relating to ballistic electrometers and electrostatic 

 Siemens dynanometers, and pointed out the possibility of 

 instruments such as he had exhibited being of use in elucidating 

 the obscure points in connection with so-called "contact electri- 

 city." The President complimented Mr. Boys on the beautifully 

 simple and remarkably sensitive electrometers exhibited. He 

 remembered that some years ago Mr. Gordon made a very 

 small electrometer, but its insulation was insufficient for electro- 

 static work. He agreed with Mr. Boys as to the advantages of 

 small instruments, providing sparking across or tilting of the 

 needle could be prevented. On the other hand, he thought the 

 use of small potential differences on the needle was a step in the 

 wrong direction, when great sensibility was required. Prof. 

 Perry asked if the needle could not be kept charged by 

 occasional contacts with a charged acid cup. Mr. Boys said he 

 had originally intended using a fairly highly charged needle, but 

 had not yet done so. He also suggested that an electrometer of 

 very small capacity might be made by reducing the quadrants 

 surrounding a disk-needle, until they became like small tuning- 

 forks.— A paper on electrification due to the contact of gases 

 with liquids, by Mr. J. Enright, and one on the expansion of 

 chlorine by heat, by Dr. Arthur Richardson, were taken as 

 read. 



Entomological Society, July i.— Mr. Frederick DuCane- 

 Godman, F.R.S., President, in the chair.— Mr. Jacoby exhibited 

 a specimen of a species of Coleoptera belonging to the family 

 Galeruada, with the maxillary palpi extraordinarily developed 

 -Canon Fowler, on behalf of Mr. Wroughton, Conservator of 

 Forests, Poona, exhibited specimens of a bug imitating an ant, 

 Folyrachis spmiger, and of a spider imitating a species of 

 Mutilla, and read the following notes :— "I have taken a good 

 many specimens of a bug which has achieved a very fair 

 imitation of Folyrachis spiniger (under the same stone with 

 which It may be found), even to the extent of evolving a pedicle 

 and spines in what, were it an ant, would be its metanotum 

 Curiously enough, however, these spines are apparently not 

 alike in any two specimens. Is it that this bug is still waiting 

 for one of its race to accidentally sport spines more like those of 



