Oct. 31, 1889] 



NATURE 



657 



and forth movements, constituting a greater part of the shaking 

 that is sensible, have usually a period of from three to five per 

 second. At the end of a disturbance the wave-period is almost 

 always very much greater than it is at the commencement or 

 middle of a disturbance. Quite recently an earthquake was 

 recorded in Japan, having a period of from six to eight seconds. 

 Observations like this are undoubtedly valuable from a scien- 

 tific standpoint, but many observations are of practical import- 

 ance. For example, we now know that a seismic survey can 

 be made of any given piece of property, and as the result of 

 such a survey it may be found that buildings erected on one 

 side of the given area may suffer very much more than those 

 upon the other side. Again, we know that, because in severe 

 earthquakes the motion at the bottom of a comparatively shallow 

 pit is very much less than it is upon the surface, buildings may 

 be partially cut off from earthquake motion by giving them 

 proper foundations. 



In addition to the theoretical and practical results which have 

 flowed from the study of earthquake diagi-ams, mechanical 

 science has gained something from the new types of machines 

 that have been evolved. Now we know how to make pendu- 

 lums astatic. New combinations in clockwork have been in- 

 vented, new governors for machinery designed, some of which 

 have already proved themselves useful for other purposes than 

 earthquake machinery. One machine, which has been the out- 

 come of seismological work in Japan, and which promises to be 

 of practical value outside the domain of seismology, is an in- 

 strument which records the vibrations in a railway train. With 

 diagrams of these vibrations before us, we are enabled to time 

 a train between stations; to see when it went quickly and when 

 it went slowly, to note the duration of stoppages, to detect 

 irregularities on a line, as, for instance, those which may occur 

 at crossings and points, those due to a want of ballast, variations 

 in gauge, imperfections in bridges, &c. 



Seismometrical observations have not alone been confined to 

 the observation of earthquakes proper, but observations have 

 been made upon disturbances produced by explosions of sub- 

 stances like dynamite and gunpowder, the falling of heavy 

 weights, and the moving of trains, &c. The records obtained 

 from these experiments have perhaps taught us more about the 

 nature of earth vibrations than we have obtained from the ob- 

 servation of actual earthquakes. So far as surface disturbances 

 in superficial soil are concerned, we now know that the pheno- 

 mena they present are anything but what we might anticipate 

 as likely to occur in a theoretically elastic material. 



An extremely practical subject which has received attention 

 in Japan has been observations and experiments upon the effects 

 produced by earthquakes upon buildings ; an account of them 

 — many of which have been successfully put into practice by 

 builders — is now being published as the fourteenth volume of 

 the Seismological Society's Transactions. As this particular 

 subject is of such vital importance to the Government of Japan, 

 who are continually erecting European structures, it is at the 

 present time being discussed by a Committee of engineers, 

 architects, and others, summoned for the purpose by the late 

 Viscount Mori, Minister of Education. 



The Imperial Government of Japan, by establishing in the 

 Meteorological Depar-tment an Earthquake Bureau, and in the 

 Imperial University a Chair of Seismology, have given an 

 impetus to seismological investigation in general. In several 

 parts of the country seismographs have been established, and, 

 at between 600 and 700 stations, records are kept of all disturb- 

 ances which are felt. From these records we are now able to 

 study the distribution of seismic activity, both with regard to 

 space and time. For instance, we now know that the greater 

 number of shocks originate on or near the eastern coast ; we 

 know that there are many "centrums," from some of which 

 weak, and from others strong, shocks originate ; that, on the 

 average, we feel in Japan at least two shocks per day. 



Inasmuch as earthquake disturbances are relatively superficial, 

 we may consider the area of disturbance as a very fair estimate 

 of seismic effort. In 1885 the land area shaken was about 

 660,000 square miles, and in 1886 it was about 562,000 square 

 miles. 



Here we have the commencement of a series of interesting 

 figures which may, perhaps, be related to a heat-gradient — to 

 the fluctuations in the flow of the Black Stream, or to something 

 not yet thought about. 



Hitherto, when studying earthquakes in relation to meteoro- 



logical changes, the position of the moon, the seasons, &c., we 

 have been compelled to take very imperfect catalogues of shocks 

 which have originated from "centrums " as independent of each 

 other as most volcanoes are of each other. Now we are getting 

 material which will enable us to study a group of earthquakes 

 which have come from a given origin. Disturbances in the 

 ocean have not been overlooked, and the waves which Japan 

 has sent to America, and those which America has returned, 

 have been carefully investigated, and average depths of the 

 ocean along several lines have been determined. In all in- 

 stances it appears that the depths calculated from the transmis- 

 sion of a sea-wave are a little less than the depths obtained by 

 averaging the soundings. Does this mean that there is an in- 

 completeness in the formulae which have been used ? or does it 

 mean that sailors have allowed their line to run a little after 

 striking bottom ? 



Other investigations which have been made relate to the 

 effects produced by earthquakes upon the lower animals ; and 

 one investigator, Prof. Sekiya, at one time kept pheasants pur- 

 posely to observe their behaviour at the time of an earthquake. 

 A conclusion arrived at is that pheasants, geese, horses, and 

 other animals often show decided symptoms of alarm a few 

 seconds before the occurrence of a severe shaking. The reason 

 of this probably is that they are more sensitive to preliminary 

 tremors than human beings. 



The relationship between volcanic phenomena, earth currents 

 and magnetism, and earth disturbances, has not been unnoticed^ 

 while mathematicians and physicists have had new problems 

 suggested to them respecting the determination of earthquake 

 origins, the depth of "centrums," the force required to cause 

 destruction like the shattering and overturning of structures, the 

 propagation of surface waves, &c. 



The observations of late years respecting the destruction of 

 submarine cables have led Mr. Forster, of Zante, to the opinion 

 that certain earthquakes are the immediate result of submarine 

 landslips, and suggested to Japanese observers that something 

 might be learnt by periodical soundings made along the Japanese 

 shores. Volcanoes have not been overlooked, and many new 

 facts have been obtained for vulcanologists. For instance, we 

 now know that many volcanoes have a definite curvature de- 

 pendent upon the density and strength of the materials of which 

 they are built, and given any two of these three quantities— curva- 

 ture, density, and strength — we may determine the third ; thus, 

 as pointed out by Mr. Becker, of the United States Geological 

 Survey, may not the shape of lunar volcanoes, with an assump- 

 tion as to the density of the material of which they are composed, 

 lead to an opinion as to the materials out of which they are 

 made? 



About the extremely minute movements called earth tremors, 

 which are probably in all places and at all times to be observed, 

 much has been done. For three or four years, by a specially- 

 contrived instrument, these have been recorded automatically. 

 The investigation of these records has led to the conclusion 

 that earth tremors are closely connected with wind. W^hen a 

 heavy wind is blowing, tremors are usually strongly marked, but 

 the more curious result is that they are often very marked in 

 calm weather. An inspection of the tri-daily weather maps 

 published in Japan shows that on these occasions a heavy wind 

 is blowing against high mountains 60 to 200 miles distant. 

 From this it appears that earth tremors outrace storms inland, 

 much in the same way that small waves outrace storms upon the 

 ocean. Insomuch as earth tremors accompany heavy baro- 

 metrical gradients, and these are related to the outpouring of 

 fire-damp in our mines, it would appear a legitimate investiga- 

 tion to study the behaviour of a tromometer, say in the Newcastle 

 area, in relation to the escape of underground gas. Hitherto I 

 believe that investigators in Great Britain have been observing 

 seismometers rather than tromometers. 



With these few remarks respecting the general nature of 

 investigations which have been made and are yet going on in 

 Japan, I will enumerate a few phenomena which require ex- 

 planation, and suggest a few investigations which have yet to be 

 undertaken. 



Large earthquakes are usually preceded by a series of short - 

 period vibrations. These vibrations have an amplitude of about 

 one-tenth of a millimetre, and six to ten of them occur per second. 

 With a seismograph giving great multiplication, it is probable 

 that still smaller and more rapidly recurring waves may be 

 recorded. These hitherto unseen portions of an earthquake 



