September 6, 1894] 



NA TURE 



451 



Potsdam (Dr. Eschenhagen). — First shock at loh. 34111. 44s. 

 a.ni., amplitude \' ; second shoclc at loh. 36111. 24s., amplitude 

 •9-10' ; third shocl: (slight) at loh. 41m. 14s. ; all three 

 registered on the declination curve. 



Wilhdmshaven (Or. C. Burgen).— The needles of two in- 

 struments (declin.iiion and Lloyd's balance) oscillated so 

 strongly that the light-point for some time made no impression 

 on the paper. Declination : from loh. 37m. 55s. to 

 loh. 43m. 25s. a.m., gap in the curve, oscillations perceptible 

 till loh. 49m. 55s. ; the needle then became steady, but oscilla- 

 tions were again visible from loh. 52m. 255. to loh. 56m. 25^-. 

 Bifilar : magnet in oscillation from loh. 3Sm. 55s. to 

 loh. 44m. 55s Lloyd's balance : from loh. 3Sm. 25s. to 

 foh. 47m. 55s., gap in the curve, then till loh. 54m. 25=. 

 oscillations traceable. 



Utrecht (M. M. Snellen). — The magnetic curves show a dis- 

 turbance beginning at loh. 37m. a.m., reaching a maximum at 

 loh. 41m., and followed by a second maximum at loh. 52m. 46^. 



Pare Saint- Maur (iM. Moureaux). — The pulsations are regis- 

 tered on the magnetic curves at loh. 40m. a.m. They are 

 most marked on the hifilar curve, and much less appreciable on 

 the vertical force curve. The declination curve shows a second 

 movement seven or eight minutes after the first. The gravity- 

 barometer curve indicates a slight depression of the mercury 

 at the moment of the principal shock. The two copper bars, 

 with bifilar suspension, orientated north-south and east-west, 

 were undisturbed by the shock.' 



Kevj (Mr. C. Chree). — There is a very slight movement on 

 both the horizontal force and declination curves, from loh. 41m. 

 to loh. 46m., that on the former being the more conspicuous. 

 I'he vertical force curve shows no disturbance. 



In the following table are given the velocities of the eatth- 

 quake-pulsaiions obtained from the different records. The 

 epoch used in each case is the beginning of the pulsations, and 

 the initial time that given by M. Coumbary. 



The average velocity of the pulsations of the principal earth- 

 quake obtained from the above ten records is 3^42 i '12 km. 

 per sec.'- A more reliable estimate will probably be obtained 

 by omitting the records from Nicolaiew and Utrecht ; the 

 former because the horizontal pendulum is far more sensitive 

 than the magnetographs, and the latter because the time given 

 IS that of the beginning of the small pulsations. The remaining 

 records (of the beginning of the large pulsations) give for the 

 average velocity 3*20 ± 07 km. per. sec. 



This value agrees very closely with that obtained for the 

 pulsations of the Greek earthquake of April 27, 1S94, namely 

 3"2i ± '07 km. per second (Report of the B.A. Earth Tremor 

 Committee, 1894), and also with several values determined by 

 Dr. von Rebeur-l'aschwilz. It hardly differs, again, from the 

 average velocity between Constantinople and Bucharest, showing 

 that the pulsations and the vibrations which constitute the shock, 

 even if they were independent, travelled at about the same rate. 



There is some doubt as to the identity of the second group of 

 pulsation^ recorded on some of the curves with one another or 



1 It h:is been inferred, from the steadines.s of the copper bars, that the 

 Tiiovenient of the needles is of magnetic and not of mechanical origin (sec 

 ' '^mptes RciuinSy vol- cxix- 1894, pp._ 251-252, and N.\tuke, vol. 1. 1894, 

 ('. 394). But is not the centre of gravity of the copper bar equidistant from 

 the two points of support, and tliat of the declination needle nearer the south 

 support (in order that the needle may rest horizontally)? And, if so. might 

 not the movement be of mechanical origin ? (See Geot. Mag. vol. ii. 1885, 



pp- 2IO-2It). 



- With the initial times as given by the two stopped clocks mentioned by 

 Mr. Wrench, the corresponding values of the velocity would be 2"30 and 

 2 "56 km. per sec. 



with one of the earthquakes at Constanlinople. Those at Wil- 

 helmshaven and Utrecht seem to follow by too long an interval 

 to be due to the third shock at loh. 35m. a.m. G.M.T. It 

 is not impossible, however, that the second group at Pola, 

 Potsdam, and Pare Saint-Maur may be connected with this 

 earthquake. If this be the case, the velocity of the pulsations 

 to these places would be 3-28, 479, and 299 km. per second, 

 respectively, and the average of the three 3 69 km. per second. 



Charles Davison. 

 King Edward's High School, Birmingham. 



Photo-electric Phenomena. 



A SHORT time ago, a report was given in N.\TURE (vol. xlix. 

 p. 226) of modern researches respecting the photo-electric dis- 

 charge of negatively electrified bodies. But no mention was 

 made of our investigations on the same subject, although they 

 were published in a series of articles, from 1889 to 1894, in 

 Wiedemann' s Annalen and in the IViener Berichte. ^.% the 

 results of these researches seem to be unknown to English 

 physicists, we enumerate them here in brief, in the hope that 

 they will prove of some interest. 



Well -cleaned plates of aluminium, magnesium, and pure or 

 amalgamated zinc are,when negatively electrified, discharged in a 

 few Seconds by the light of the sun and the cloudless sky. The active 

 rays extend from the blue part of the spectrum to the farthest ultra- 

 violet, so that the active r.ays are almost completely absorbed 

 by transmission through glass. Hence a well-amalgamated 

 ball of zinc connected with Exner's aluminium-leaf electro- 

 scope may be used as a photometric apparatus to settle the 

 amount of ultra-violet radiation emitted by the sun and the sky 

 which reaches the earth's surface. Measurements of this kind 

 were made by us in 1890, during June in Wolfenbuttel, and 

 during July on the top of the Sonnblick. The discharging 

 power of sunlight at a level of 3100 metres (i.e. on the Sonn- 

 blick) was found to be twice as great as at a level of So metres 

 (Wolfenbuttel), corresponding to the greater proportion of 

 blue and ultra-violet rays. These results are in conformity 

 with Langley's well-known researches on the absorption of the 

 blue sunlight by the atmosphere. 



The more electro-positive a metal is, the larger the wave- 

 length of light capable of producing a photo-electric discharge. 

 Photo-electric cells of high sensibility were therefore made by 

 using cathodes of metallic sodium, potassium, and rubidium, 

 I arranged in a glass bulb, and dipping in an atmosphere of 

 rarefied hydrogen. The leakage of negative electricity from a 

 sodium or potassium surface is produced even by the light of a 

 candle at six or seven metres distance from the cell, and 

 rubidium-cells are sensible to the weak light sent out from a 

 red -hot glass-rod. 



The photoelectric discharge by the action of ordinary day- 

 light is also shown in a clear way by some non-metallic bodies, 

 e.g. by the phosphorescent combinations of calcium with 

 sulphur, and by the dark-coloured kinds of fluorite. Other 

 minerals show traces of the same phenomenon. It is, there- 

 fore, probable that the sunlight and the daylight cause the 

 negative electricity of the earth's surface to be partly dissipated 

 into the air. If this theory were correct, it would give a 

 foundation to the explanation of the daily and yearly variation 

 of atmospheric electricity. For three years we have together 

 made measurements respecting the ultra-violet radiation of 

 the sun and the amount of atmospheric potential at the same 

 time, and the results agree with this supposition. 



By using the liquid alloy of potassium and sodium, which, 

 in an atmosphere of rarefied hydrogen, shows a reflecting 

 surface, we are able to study the influence of polarised light 

 upon the photo electric discharge. The maximum intensity of 

 the photo-electric current is observed when the plane of polarisa- 

 tion is perpendicular to th.at of incidence. 



The photo-electric discharge is sensitive lo exterior magnetic 

 influences in a similar way to the luminous discharge in a 

 Geissler's tube. In a strong magnetic field it almost entirely 

 ceases. 



Electrical vibrations set up by the Hertz vibrator pass through 

 a Geissler tube provided with an alkaline metal electrode far 

 easier in daylight than in darkness. With a sensitive arrange- 

 ment the weakest traces of light are sullicient to start the 

 luminous discharge. 



In one case the disruptive electrical discharge ceases when 

 light is applied. If the sparks of an influence machine are 



NO. 1297, VOL. 50] 



