550 



NA rURE 



[ArKii, lo, 1902 



perofskite in the more basic nephelinites, and as sphene in the 

 phonolites which are of the more ordinary type without soda- 

 amphiboles. Most of the specimens from ^It. Elgon and the 

 neighbourhood consist of nephelinites, but in some of them 

 the nephcline, both as phenocrysts and in the groundmass, is 

 partially or wholly replaced by melilile. Examples of melanite- 

 nepheline rocks allied to borolanile, and of monchiquite dyke- 

 rocks from Mt. Elgon, were also described. A specimen of 

 nephelinite from the neighbourhood of Ruwenzori containing 

 much perofskite suggested ihe contemporaneity of the eruptive 

 rocks of Mt. Elgon and of the volcanic region at the foot of Mt. 

 Ruwenzori. 



Royal Meteorological Society, March 19. — Mr. W. II. 

 Dines, preiideni, in the chair. — Mr. VV. N. Shaw, E. R.S., read 

 a paper on la lune mange les nuages, which was really a note 

 on the thermal relations of floating clouds. He also exhibited 

 an arrangement of apparatus whereby the conditions applicable 

 in the case of a floating cloud can be experimentally realised. — 

 Mr. F. J. Brodie read a paper on the prevalence of gales on the 

 coasts of the British Islands during the thirty years 1871-1900. 

 The total number of gales of all kinds dealt with during this 

 period was J455, the yearly average being 485, of which io'6 

 were severe. The worst year was 1S83, while the quietest was 

 18S9. The stormiest month was January, 1890. At all seasons 

 of the year excepting the summer, the prevalence of gales from 

 south-west is greater than from any other quarter. The 

 minimum of such gales is reached in the spring, when rather 

 less than 20 per cent, are from south-west, more than half the 

 storms being, however, from points between south-west and 

 north-west. The prevalence of gales from polar directions is 

 then at its maximum, more than 21 per cent, blowing from 

 points between north and east. In the spring of 1S83, out of a 

 total of eleven gales no fewer than seven were from these 

 quarters, the proportion being about three times the average. 

 The highest velocities recorded were those at Fleetwood during 

 the westerly gales on December 22, 1S94, and on January 12, 

 1899. On the former. occasion, for nine hours, from 7 a.m. to 

 3 p m. , the mean velocity was sixty-four miles per hour, and at 

 9 a.m. it reached a maximum of seventy-eight miles. It appears 

 that on the average 43 per cent, of the storm systems which visit 

 our coasts advance from some point of the compass lying between 

 .south and south-west, and travel towards some point lying 

 between north and north-east. 39 per cent, have an easterly 

 motion, while less than I per cent, move westwards. A mean 

 of 264 cases shows th.at the deep cyclonic systems which visit 

 our islands travel on an average at the rate of 24"i miles per 

 hour ; in some cases, however, the rate was not more than eight 

 or ten miles, while in others it amounted to forty, fifty and even 

 sixty miles per hour. The author concluded his paper by 

 exhibiting a series of weather maps showing the progress of 

 some of the most notable gales during the period covered by the 

 discussion. 



CAMnRincE. 



Philosophical Society, March 3 — Prof. Macalister, presi- 

 dent, in the chair. — On a method of increasing the sensitiveness 

 of Michelson's interferometer, by Mr. H. C. Pocklington. It is 

 shown that the sensitiveness of Michelson's interferometer can 

 be greatly increased if we can cause the interfering beams to 

 be circularly polarised in opposite senses. This can be done 

 appro.ximately in the ordinary form of the instrument (with, 

 however, an unsilvered inclined mirror) by placing an eighth-wave 

 plate of mica in front of each of the perpendicular mirrors so 

 that a principal axis of each plate is parallel to the line of inter- 

 section of the mirrors, and suitably choosing the plane of 

 polarisation of the incident light. A theoretically better method, 

 in which the inclined mirror and the compensating plate are placed 

 rather less than a quarter-wave apart, is discussed, but it is con- 

 cluded that only experiment can decide which will be the better 

 method in practice. — The influence ofcurrents in metals on reflected 

 and transmitted light, by Mr. P. V. Bevan. — (a) On the conduc- 

 tivity of the vapours of the alkali metals ; (/') on induced radio- 

 activity, by Prof. J. J. Thomson. The investigation was under- 

 taken with the intention of seeing whether the " induced radio- 

 activity '' shown by a metal rod after long-continued negative 

 electrification in the open air would occur if the rod were placed 

 in a closed vessel instead of outside in the open air. The closed 

 vessel was a zinc gasometer 102 cm. high and 75 cm. in 

 diameter ; the vessel was insulated and used as one of the 

 electrodes, the other electrode was a metal tube placed at the 



NO. 1693, VOL. 65] 



axis of the cylindrical gasometer. A potential difference of 800 

 volts between the cylinder and this rod was produced and the 

 current between these electrodes was measured. This current 

 was " saturated " and was therefore a measure of the total 

 ionisation in the gas in the vessel ; if the rod became radio- 

 active, the ionisation and therefore the current would increase. 

 The current was measured in the morning, and the rod in the 

 vessel kept connected with the negative terminal of a Wimshurst 

 machine for six or seven hours, when it was disconnected from 

 the machine and the current again measured ; if the gas in the 

 vessel were not exposed to Rontgen rays whilst the rod was 

 negatively electrified, the author was not able to detect any 

 increase in the current through the gas as the result of the long 

 negative electrification : if, however, the gas were exposed lo 

 Rontgen rays during the negative electrification of the rod, then 

 a well-marked increase in the current took place — the increase 

 being some t6 or 17 per cent. ; this increase was due to 

 some alteration in the rod and not to a change in the gas 

 in the vessel, for if a rod similar to the one which had 

 been electrified, but which had not itself been electrified, 

 were substituted, the current sank to its former value. No 

 increase took place in the current if the rod were positively 

 electrified. A number of experiments were made on the currents 

 through the vessel when the vessel was not exposed to rays and 

 when the rod was not electrified. Rods of different sizes and 

 different metals were tried — these all gave approximately the 

 same current ; if the rod were carefully wrapped round with dry 

 filter paper, the current showed a decided increase, while if the 

 filter paper were damp, the current was many times its value for 

 the bare rod ; the current in this case is greatest when the 

 negative ions move up to the paper-covered rod — a large effect 

 is also produced when the paper is wetted with brine or alcohol, 

 but a .solution of H.,0., produces by far the largest effect yet 

 found. — On the Hail effect in gases at low pressures (second 

 paper), by Mr. H. A. Wilson The experiments described in this 

 paper are a continuation of those described in the paper entitled 

 " On the Hall Effect in Gases at Low Pressures" (Proc. Camb. 

 Phil. Soc , vol. xi. pt. iv. ), read to this .Society last October. 

 Measurements have been made of the Hall effect and electric 

 intensity in the uniform positive columns in oxygen and hydrogen, 

 and also of the variation of the Hall effect along the discharge in 

 air at various pressures. The Hall effect in hydrogen is found 

 to be capable of being represented by the equation 



s-.= 2-65 X 10 - — , 

 / 

 where c is the transverse electric intensity or Hall effect, H the 

 magnetic field and / the pressure in millimetres of mercury. 

 The corresponding equation for oxygen is found to be 



£ = 3-Sxio-3li. 



The electric intensities in hydrogen and oxygen are found lo be 

 given by the equations 



X=2S^'/ 



and X = 26'9 ^p respectively. 



The results on. the variation of the Hall effect along the dis- 

 charge are shown by curves which resemble the curves showing 

 the variation of the electric intensity along the discharge. The 

 pa|ier concludes with some theoretical discussicn of the results. — 

 On the extraction of the gases from one cubic centimetre of 

 blood, by Mr. J. Barcroft. — On the coefficient of mutual induc- 

 tion between a circle and a circuit with two parallel sides of 

 infinite length, by Mr. G. F. C. Searle. — Notes on Semper'^ 

 larva, by Mr. K. Ramunni Menon. 



Edinburgh. 

 Royal Soci"ty, February 17. — Prof. Geikie in the chair. 

 — Major W. B. Bannerman, superintendent of the Plague 

 Research Laboratory, Parel, India, gave an account of the work 

 carried on in the laboratory of which Mr. Ilaflkine is director 

 in chief. A staff of fifty-three, of whom five are European, now 

 occupies a huge building which has been in succession a Roman 

 Catholic seminary, a Governor's residence, and a plague 

 hospital. A detailed account was given of the preparation 

 of Maffkine's prophylactic, which the Indian Clovernment 

 had adopted as a vaccine for the plague. Some interesting 

 statistics were given showing that under precisely similar con- 



