July 28, 1904] 



NA TURE 



309 



to October 30, but, as the comet will be only 03 of its 

 original magnitude on July 30, and is still decreasing, it 

 is scarcely worth while reproducing it here. 



The R.'A. is varying but little, and on August i will be 

 i2h. i6m. 40S., whilst the declination is slowly decreasing, 

 and on the same date will be +47° 34'-6. 



A Modified Form of the Newtonian Reflector. — In 

 the Monthly Notices for May, 1895, the Rev. Chas. Davies 

 described a modified 'orm of Newtonian reflector in which 

 the rays from celestial objects fell on a large plane mirror 

 fixed at the open end of a horizontal tube, and were thereby 

 reflected on to an ordinary parabolic mirror fixed at the 

 olher end, afrerwards being brought to a focus through an 

 aperture in the centre of the plane mirror to which was 

 affixed the observing eye-piece. A movement of the plane 

 mirror about the optical axis, and of the horizontal tube in 

 azimuth, allowed any point in the sky to be reached. 



By fixing the telescope in a fork at the upper end of a 

 polar axis, M. E. Schaer, of Geneva, now proposes to 

 modify this instrument so that, whilst retaining its original 

 advantages, such as the unchanging position of the eye- 

 piece, it may be used like an ordinary equatorial and 

 coelostat, and by the simple rotation of the polar axis by 

 clockwork the object may be kept stationary in the centre 

 of the field. In this arrangement the mirrors are so placed 

 that they suffer very little from flexure caused by changes 

 of position, tsing a model instrument constructed on these 

 lines M. Schaer found that the practical results were excel- 

 lent (Astrononiische Nachrichten, No. 3<)s8). 



SEISMOLOGICAL NOTES. 



"T^HE sixteenth number of the Puhlications of the Earth- 

 quake Investigation Committee in Foreign Languages 

 (Tokyo) consists of 117 quarto pages of print and 9 full-page 

 illustrations. The subject is on Milne horizontal pendulum 

 seismograms obtained at Tokyo, the author of which is 

 A. Imamura, assistant professor of seismology at the 

 Imperial University of Tokyo. 



While discussing amplitudes, it is pointed out that these 

 quantities may be increased or decreased according to the 

 relationship existing between the periods of earth move- 

 ments and the period given to the pendulum, an objec- 

 tion, as has frequently been pointed out, to pendular 

 apparatus in general. Out of a list of 298 records (July 24, 

 1S99, to December 24, 1902), the more important are con- 

 sidered in relation to corresponding records obtained from 

 other types of instruments in Japan, and from similar types 

 of instruments in various parts of the world, the registers 

 from which are issued biannually by the British Associ- 

 ation. The more important results relate to the speeds with 

 which different phases of earthquake motion have been 

 propagated over paths of great length. By means of more 

 than forty diagrams, each referring to a particular earth- 

 quake, speeds along arcual paths for several of the more 

 important phases of motion are represented by straight 

 lines, that is to say, the speeds are constant. For certain 

 disturbances the evidence leads us to this conclusion, but 

 this is not the case for all. For example, in Fig. 4, a 

 diagram similar to publications by the British Association 

 (Report, 1902, p. 66), we notice in connection with the 

 preliminary tremors that the longer the wave path the 

 greater are the divergencies among the observations which 

 give the time interval to traverse the same. The time taken 

 to travel 25° has apparently varied between 35 and 4 

 minutes, that is to say, the observations agree within 

 30 seconds. For So°. however, the divergence is 5 minutes, 

 while on still longer paths the intervals are still greater. 

 When we look at these variations as shown on squared 

 paper, we should certainly hesitate before representing their 

 mean position by a straight line. If, however, it is a 

 straight line, and we know the recording instruments to 

 be similar, then one inference is that minute tremors which 

 may be recorded at a station near to an origin may have 

 failed to reach or to make themselves evident at stations 

 which are very remote. 



But why should earthquake vibrations fall in line with 

 the vibrations of elastic bodies? If our world has a fluid 

 or gaseous nucleus, .\rrhenius, Fisher, and other physicists 



NO. 1813, VOL. 70] 



and geologists see in the same an explanation for many 

 phenomena. Convection currents might explain slight 

 changes in latitude (Fisher), and they certainly suggest 

 variability of velocity along the same path. 



Although we do not agree with all Mr. Imamura's con- 

 clusions, seismologists are indebted to him for a piece of 

 valuable research. 



In vol. ii.. No. 6, of the reports of the Tokio Physico- 

 mathematical Society Mr. K. Honda gives an account of the 

 daily periodic changes in the level in an artesian well the 

 depth of which is 380 m. with a water head within 32 m. 

 of the surface of the ground. What he found was that 

 there were two maxima and minima every twenty-four 

 hours, the range of motion varying between 3 cm. and 

 I cm. (For somewhat similar experiments made in a 

 shallow well close to the bore-hole here considered, see 

 Reports Brit. .'Xssoc, 1895, p. 104.) 



Near to the days of full and new moon the movements 

 are marked and regular ; the phases of maxima and minima 

 agree with those of the tides in Tokyo Bay. The well 

 sinks with a high barometer and rises with a low baro- 

 meter. Rain does not affect the level. By experiment it 

 was found that variation of pressure of i mm. of mercury 

 produced a change in the level of the water of 13-5 mm. 

 -An equal natural pressure acting on the water head causing 

 it to sink, and on the surrounding ground causing it to 

 rise, only results in a level change of 435 mm. From this 

 it is concluded that the earth's crust only transmits 68 per 

 cent, of pressure on its surface to a depth of 380 m. Another 

 conclusion is that the daily fluctuation of i to 3 cm. is 

 more likely to be a tidal than a barometrical effect. The 

 distance to the sea is 3 km. In a deep well in Yokohama 

 0-6 km. from the sea, the tidal effect results in a change 

 of level of 16 cm. This extremely interesting paper con- 

 cludes with references to the frequency of earthquakes in 

 relationship to fluctuation in barometric and tidal loads. 

 In No. 9 of the same reports Mr. Honda gives a continu- 

 ation of similar researches carried on at three other deep 

 wells, at the end of which he shows that earthquakes with 

 a submarine origin are most frequent when tidal 

 pressure is at maximum, a minimum, and when the rate 

 of pressure is changing most rapidly. No. 8 of the Journal 

 is from the pen of Dr. F. Omori, who shows, chiefly from 

 the consideration of after shocks, that earthquake frequency 

 is affected bv changes in atmospheric pressure. 



Consul G.' Para, of Uskub, gives (Kaiserliche Akad. d. 

 Wissetischaften in Wien, April 21, No. 10) a few statistics 

 relating to the destruction caused by the earthquake which 

 on .April 4 disturbed the Balkans. This is followed by 

 further details of a more geological character by Prof. R. 

 Hoernes. The phenomena described are of an ordinary 

 character. 



Under the title of " L'Eruzione dell' Etna in 1892," 

 (vol. i.), in a large quarto volume, the director of the 

 observatory in Catania, Prof. A. Ricco, and S. Arcidiacono 

 give a detailed account of the phenomena which accom- 

 panied the eruption of Etna in 1892. As an assistance to 

 the better understanding of the historical sequence in 

 events, this is prefaced by accounts of the eruptions of 

 18S3 and 1896, all of which took place on the line of a 

 radial fracture at points from 1000 to 1500 m. lower than 

 the main central crater. It is essentially a volume of 

 observations of value to the vulcanologist, to be followed 

 at a later date by deductions. 



In the Bollettino dell' Accad. Gioenia, Catania, fas. Ixxix., 

 December, 1903, S. Arcidiacono gives a short account of 

 earthquakes which recently disturbed Etna, and which were 

 of local origin. From a tabular statement of these it 

 appears that from i8g8 seismic activity was fairly uniform 

 and not pronounced, but after the eruption of 1902 it became 

 three-fold. 



The first paper in the Bollettino della Societa Sismologica 

 Italiana (vol. ix., No. 9, 1903-4), by M. Tito Alippi, relates 

 to the possible relationship of bonniti and bombiti 

 (mist poeffers, barisal guns, &c.) to seismic movements. 

 From a list of seismic disturbances recorded in a district 

 where bonniti were frequent, it does not appear that the 

 two phenomena are connected. The multiplication of the 

 seismograph was, however, only 12. Had it possessed 

 ten times this sensibility it might have responded to minute 



