3IO 



NATURE 



\yan. 30, 1890 



the figures are obtained, and has arrived at the following con- 

 clusions : (i) the nature of the dielectric plate does not change 

 the character of the figures produced, and (2) the nature of the 

 powders used seems to have no material effect on their shape. 

 In the course of his experiments he has found a mixture of sub- 

 limed sulphur and lycopodium to give better figures than the red 

 lead and sulphur usually employed, and also that a large and 

 highly polished knob is advantageous, particularly when the 

 Leyden jar is charged negatively. Sometimes when obtaining 

 negative figures, nebulous patches occur, and these were attri- 

 buted to the so-called electric winds sent off from roughnesses on 

 the knob when not sufficiently well polished. If instead of bring- 

 ing the knob in contact with the plate, it is only brought near to 

 it, then a peculiar figure closely resembling a "splash" results. 

 A positive splash consists of short lines radiating from the point 

 -of approach, whilst a negative splash is made up of more or less 

 rounded spots which become elongated in a radial direction as 

 their distance from the centre of the splash increases. Negative 

 splashes are, however, much more difficult to produce than posi- 

 tive ones. When viewed in the dark, the discharge producing 

 the splash is seen to consist of a bundle of small sparks which 

 branch outwards on approaching the plate. In conclusion the 

 author remarked that roughnesses on a conductor produced more 

 electric winds when the conductor is charged negatively than 

 when positively charged, and invited the opinions of members as 

 to the causes of the differences observed between positive and 

 negative electricity. Prof. RUcker said he had recently obtained 

 figures produced by discharges on photographic plates. Generally 

 he observed that negative discharges produce roundish patches, 

 whilst positive ones give more filamentary figures. On passing 

 a spark across a glass plate covered with lampblack, its trace 

 was found to have a black core at one end, whilst the other was 

 quite clear. He also made remarks on the distinctive character 

 of the positive and negative discharges in partial vacuo, and con- 

 sidei'ed investigations as to the causes of such differences to be of 

 great importance. Prof. Adams thought any attempt to discover 

 the causes of such differences as those noted in the paper was to 

 be commended, for the well-known fact that it is more difficult 

 to insulate a negative charge than a positive one has long needed 

 an explanation. — A paper on galvanometers, by Prof W. E. 

 Ayrton, F. R.S., T. Mather, and W. E. Sumpner, was read 

 by Pi-of. Ayrton. In fitting up the Physical Laboratories of the 

 Central Institution of the City and Guilds of London Institute, 

 the authors have had occasion to obtain galvanometers of various 

 types and patterns, some of which have been made to special 

 designs, and specimens of instruments embodying recent im- 

 provements were exhibited at the meeting. The question as to 

 whether fairly sensitive galvanometers should be astatic or non- 

 astatic was answered in favour of the former system, from the 

 fact of its being less affected by external magnetic disturbances, 

 and the greater ease with which great sensibility may be obtained. 

 The usual method of placing the mirror inside the coil was shown 

 to be undesirable, and in the newer forms of instruments Mud- 

 ford's improvement of placing the mirror outside the coils has 

 been adopted ; the space near the axis of the coil being nearly 

 filled with wire. It was also shown that if wire be wound in a 

 certain approximately spheroidal space near the magnets, then 

 these convolutions will tend to oppose the more distant portions of 

 the coil ; however, by winding the two parts in opposite direc- 

 tions they conspire to deflect the magnet. Details as to methods 

 •of supporting the coils were then discussed, and the importance of 

 fitting them in boxes mounted on hinges or otherwise, so as to 

 be readily removable, was pointed out. A galvanometer devised 

 for teaching purposes, and provided with variable damping 

 arrangements was described, in which the damping is effected by 

 enclosing the mirror in a glass cell whose sides can be caused to 

 approach or recede by turning a milled head outside the instru- 

 ment. This arrangement enables the damping to be varied 

 between wide limits, and its effect on the swing produced by a 

 given discharge can be determined. The instrument is also 

 serviceable both as an ordinary damped galvanometer, or as a 

 fairly ballistic one. In measuring quantities of electricity by the 

 first swing of a galvanometer needle, a correction has usually to 

 be introduced for damping ; this correcting factor is simple 

 enough when the damping is small, but becomes more complicated 

 as the damping increases, and to facilitate the calculations a table 

 of values of the factor for various values of A. (the logarithmic 

 decrement) has been calculated. From this it appears that, for 

 values of A less than o'5, the value of the factor is very nearly 

 (i -V \\), the correction usually employed. Improvements in 



methods of insulating the coils and terminals of galvanometers 

 required for insulation tests were next described, the principle of 

 which may be gathered from Figs. 107 and 108 in Prof Ayrton's 

 " Practical Electricity." A special form of instrument for high 

 insulation work was exhibited, in which the copper resistance of 

 the coils is nearly 400,000 ohms, and the shortest path along 

 which surface leakage can take place from the coils to the base of 

 the instrument is between 30 and 40 inches of ebonite artificially 

 dried by sulphuric acid. This is attained by supporting the coils 

 from two corrugated ebonite rods which depend from a brass ring 

 carried on the top of three corrugated pillars fixed to the base plate. 

 The instrument was constructed to drawing and specification by 

 Messrs. Nalder Brothers, but the method of supporting the coils 

 was 'suggested by Messrs. Eidsforth and Mudford. With re- 

 ference to the proportionality of deflection to current in reflecting 



fgalvanometers, it was pointed out that ordinary instruments 

 may differ as much as 2 per cent, within the limits of the scale, 

 hence showing the necessity for calibration when any approach to 

 accuracy is desired. Galvanometers of the D'Arsonval type some- 

 times differ from proportionality quite as much as the one above 

 referred to, but by fitting such instruments with curved pole 

 pieces, and all )wing the coil to hang freely from the top suspen- 

 sion, a proportionality true to less than o"i5 per cent, has been 

 attained over a scale about 30 inches long. Coming to the ques- 

 tion of sensitiveness, the importance of keeping the wire as close 

 as possible to the magnets was brought prominently forward, as 

 well as the necessity of reducing the " figures of merit " of various 

 instruments to the same standard, in comparing their sensibilities. 

 The standard adopted as most convenient and closely approximat- 

 ing to practical usage is arrived at by supposing the distance of 

 the mirror from the scale to be equal to 2000 scale divisions, and 

 the sensibilities for current and quantity are given as scale 



.divisions per micro- ampere, and scale divisions per micro-coulomb 

 respectively. The period of oscillation is also taken into account. 

 A table showing the resistances, sensibilities, coefficients of 

 self-induction and volumes of the coils of various instruments, 

 together with the relations existing between them, accompanies 

 the paper, and from this it appears that in the best astatic double 

 coil instruments, of from 10,000 to 30,000 ohms resistance, the 

 number of scale divisions per micro-ampere may reach 400 times 

 the resistance to the |th power (400 R^) when the period is 10 

 seconds. In obtaining data of various instruments the authors 

 have consulted, amongst others, Prof. Threlfall's paper on the 

 measurement of high specific resistances, in the Phil. Mag. "for 

 December 1889, and noticed two serious errors. The first of 

 these makes an instrument constructed according to Messrs. 

 Gray's pattern nine times less sensitive than it actually was, 

 whilst the sensibility of a form recommended in the paper is 

 given seventeen times too great. On account of the lateness of 

 the hour, the discussion was adjourned till February 6, before 

 which time it is hoped that a fairly full abstract will appear 

 in the technical papers. 



Geological Society, January 8.— W. T. Blanford, F.R.S., 

 President, in the chair. — The following communications were 

 read : — On some British Jurassic fish-remains referable to the 

 genera Eurycormus and Hypsocormus, by A. Smith Woodward. 

 Hitherto our knowledge of the Upper Jurassic fish fauna has 

 been mainly derived from specimens found in fine lithographic 

 stones, where the various elements are in a state of extreme 

 compression. Within the last few years remains of similar fish 

 have been discovered in the Oxford and Kimeridge Clays of 

 England, and these are of value for precise determination of 

 certain skeletal features in the genera to which they belong. 

 The author described Emycormus grandis from the Kimeridge 

 Clay of Ely, a large species which makes known for the first 

 time the form and proportions of several of the head-bones in 

 this genus. A technical description of all the bones the charac- 

 ters of which are distinguishable was given, and the author con- 

 cluded that there is considerable similarity between the head of 

 Eurycormus and the recent Ganoid Amia, even to minute points 

 of detail. He further described Hypsocormus tenuirostris and 

 //. Leedsii from the Oxford Clay of the neighbourhood 

 of Peterborough, the osteology of this genus not having 

 as yet been elucidated. Portions of the jaws have been dis- 

 covered, affording valuable information as to the form and 

 dentition of the principal elements. These jaws are not precisely 

 paralleled by any other Jurassic genus, though they possess a 

 resemblance to Pachycormus, as also to the Upper Cretaceous 

 genus, Protosphyi'cena. The President remarked that Amia is a 



