46 



NATURE 



[November io, 1898 



SCIENTIFIC SERIALS. 



Bulletin of the American Malhcmatical Society, October. — 

 The number opens with an account of the fifth summer meeting 

 of the Society, which was held at the Massachusetts Institute ol 

 Technolog)-, Boston. As on former occasions, the Society was 

 aftiliateci with Section A of the American .Association for the 

 Advancement of Science. There was a large attendance of 

 members, and at the three sessions which were held twenty-five 

 papers were communicated. Short abstracts of papers by 

 Messrs. Blake, Chessin, Lovett, Baker, Hall, Moore, Conant, 

 Boyd, Stecker, Dickson, Stabler and Martin are given. The 

 journals in which these papers will appear are indicated. Note 

 on the generalisation of Poincare and ("loursat's proof of a 

 theorem of Weierstrass's, by Prof. Osgood, gives an interesting 

 sketch of the theorem, accompanied by full references to original 

 memoirs. — The same author supplies a supplementary note on a 

 single-valued function with a natural boundary, whose inverse is 

 also single-valued. This is supplementary to a paper in the 

 June £»l/e/in, and states the writer's indebtedness for a simple 

 proof of his principal theorem to Prof. Hurwitz. — Prof. Chessin 

 contributes a note on the periodic developments of the equation 

 of the centre and of the logarithms of the radius vector. — The 

 theorems of oscillation of Sturm and Klein is a third paper on 

 the subject, by Prof. Bncher. Its object is to extend the results 

 previously established to some cases in which the coefficients of 

 the differential equation (discussed in the preceding papers) are 

 no longer continuous throughout the intervals considered in the 

 investigation. The article occupies pp. 22-43. There are 

 numerous vacation notes, and a long list of new publications. 

 It is worth noticing that the Society has appointed a Committee 

 to consider the question of securing improved facilities for the 

 publication of original mathematical papers (in the States). On 

 p. 48 it should be Prof. Price has resigned the Sadleian Pro- 

 fessorship (not Sadlerian, which is a Cambridge professorship). 



Symons's Monthly Meteorological Magazine, October. — Heat 

 and drought in September 1898. Reference to the regular table 

 of rainfall and temperature at fifty stations shows that the month 

 was remarkably dry over the whole of England and \Vales. At 

 twenty-one of the stations the rainfall was less than half the 

 average, and nine of them show a deficiency of 75 per cent. 

 The deficiency is most marked in the east and north-east of 

 England, where the falls were generally about a quarter of an 

 inch. September 1S95 ^'^^ very dry ; September 1898 yielded 

 a larger number of both absolute and partial droughts, but a 

 smaller number of stations with less than half the average fall. 

 To illustrate the distribution of temperature during the month, 

 a table of shade maxima at twenty-five stations is given, for 

 September 3-9 and 14-17. On all these days there are numerous 

 entries above So", while on the Sth there are eleven records of 

 90° and upwards. Although the absolute maximum at Camden 

 Square in 1898(91" 2) is unprecedented, and no other September 

 during the last forty years has had so many days above 85", yet 

 in 1865 there were two more days above 80°. — Results of 

 meteorological observations for September at Camden Square 

 for forty years, 1858-97. The average r.ainfall is 239 inches; 

 the year 1898 had only '33 inch, being the smallest September 

 fall since 185S. The mean temperature in September for the 

 forty years is 57°7. The mean for 1898 was 6i°'6, being an 

 excess of 3°'9 from the normal ; there is only one instance of a 

 higher mean, viz. 63"'6, in the year 1865. 



Wiedemann's Annalen der Pliysil; nnd Cheinie, No. 9. — 

 Electric currents produced by Ronlgen rays, by A. VVinkel- 

 niann. Like Pcrrin, the author succeeded in obtaining dift'er- 

 ences of potential between two diflerent metals under the 

 influence of v-rays, and also a steady current of about 6 x lo"" 

 amperes in a circuit containing two such tnetals. He calculates 

 an inferior limit for the ratio of the number of ionised mole- 

 cules of air traversed by. x-rays to the total number of mole- 

 cules, and finds it to be 4 6 x I0~". — Magnetic hysteresis, by V. 

 Niethammer. The loss of energy in alternate-current hysteresis 

 is greater than in magnetostatic hy.steresis. It is nearly the 

 same for sinusoidal and for fiat curves, but less for ))ointed 

 curves, for the same maximum induction. — Reflection of 

 kathode rays, by H. Starke. The rays enter a spherical vessel 

 fron. a side tube, and are reflected by a metallic mirror mounted 

 in the centre and movable by a handle outside. After reflection, 

 the rays are caught in another side lube containing an electrode 

 leading to a galvanometer. The galvanometer shows the same 

 deflection within wide limits of the orientation of the mirror. 

 Hence the deflection is diffuse, the particles being so small that 



NO. 15 I 5, VOL. 59] 



even a highly polished surface is rough 10 them. The curren 

 is greatly increased by insulating the mirror instead of connect 

 ing it to earth. The denser metals have the greater reflectivi 

 power. — Disintegration of incandescent platinum and palladiun 

 wires, by W. Stewart. The disintegration of wires rendera 

 incandescent by an electric current gradually decreases as timi 

 goes on, but is unaflected by the presence or absence of moistUB 

 in the air. When the air is exhausted, the disintegration a 

 platinum diminishes, while that of palladium increases. Plati 

 num is not disintegrated in hydrogen. — Causes of the changi 

 of resistance discovered by Branly, by D. van Gulik. Branly* 

 view that the conductivity of powders acquired under th{ 

 influence of electric radiation is due to a modification of thi 

 dielectric surrounding them is untenable, as experiments will 

 minute terminals under the microscope prove the correctness o 

 Lodge's view that the conductivity is due to mechanical contact 

 — Observations concerning coherers, by E. Dorn. To bl 

 effective, metallic powders should be somewhat easily oxidisedi 

 and should be exposed to the air so as to acquire a coating 

 oxide. Good results are obtained with iron and copper, bui 

 not with the noble metals. — -A supposed unknown constituent o 

 the atmosphere, by O. Neovius. The author finds a number 01 

 lines which aie identical in the nitrogen spectrum and the bluq 

 argon spectrum, and may be due to an unknown gas occurring 

 as an impurity in both. These lines show only a single co3 

 incidence with those of crypton, at 4736. ; 



SOCIETIES AND ACADEMIES. \ 



London. 

 Physical Society, October 28. — Mr. Shelford Bidwell, 

 F.R.S., President, in the chair. — A paper was read by Mr. W. 

 R. Pidgeon on an influence machine. The machine, which wa» 

 exhibited, consists of a pair of glass discs, rotating in opposite 

 directions upon a spindle. They are partially covered on on« 

 face with narrow radial sectors of tin-foil, each provided with a 

 small brass knob. This face of the disc, including its sector.^, 

 is then coated with insulating wax, leaving only the knobs pro-' 

 jecting through the wax. Two earthing-brushes pass through 

 two insulated fixed inductors, and .support them. The inductoi-s 

 are kept charged by exploring-points connected to each, and 

 placed so as to collect electricity from the revolving discs. B)» 

 means of brushes, the sectors on e.ach of the discs are successively 

 earthed at the moment they pass the fixed inductors, i.e. at the 

 moment that their capacity is a maximum ; and they are made 

 to deliver up their charge to the main collecting brushes at the 

 moment when they are electrically farthest from the inductors, 

 i.e. v;hen their capacity is a minimum. Hence, if there is no, 

 loss of electricity in the process, the potential at the collecting 

 brushes is proportionately high. Each sector of a particular 

 disc, as it moves away from its inductor with rising potential,' 

 induces a corresponding potential on the sectors opposed to it 

 on the second disc ; this action is cumulative. The thick coat- 

 ing of wax re-stricts leakage to the small area of the contact 

 knobs, so that surface eflects of dirt and moisture are minimised. 

 Captain J. H. Thomson, R.A., said that, apart from its 

 electrical merits, the machine possessed advantages in mechanical 

 construction. He thought there was still room for improvement 

 in this respect. The counter-shaft should be done away with, 

 and ball-bearings should be introduced. The inductor was a, 

 distinct improvc-mcnt ; he thought the efticiency might be in-. 

 creased by adding other inductors. Platinum-iridium was the 

 best material for brushes of such machines. Prof, .\yrton asked 

 what efficiency was obtained with modern influence machines,! 

 in general. Captain Thomson had founil that when run- 

 ning a machine by a motor, about So per cent, ot, 

 the power was wasted in mechanical friction ; of the 

 remaining 20 per cent, a great deal was lost as electrical 

 leakage. Prof. S. P. Thompson thought it had been pointed 

 out by Mr. Wimshurst that influence machines did not work 

 well unless there were at least two thicknesses of glass between 

 the inducing and induced conductors. That was why Mr. 

 Wimshurst put his sectors on the outer faces of the glass discs. 

 Mr. Pidgeon had departed from this. The advantage of the 

 narrow spacing of the sectors was not very apparent. Mr. 

 Wimshurst (abstract of communication) : Waxing the discs 

 reduces leakage, and increases the output ; the wax-coating 

 virtually doubles the number of plates. Inductors contribute a 

 further increase to the output. In 1S83 .Mr. Wimshurst tried 

 thick coatings of shellac, and also, duplicating the glass, with in 



