to the locality of the prime shock. They are, of course, liable to be somewhat indefinite. As an extreme, if we were to inquire within what region the recent Japanese disaster was felt by human beings, we might be confronted with exceptional cases at very great distances indeed; for there is some evidence that it was actually felt in England. A lady of peculiar sensitiveness has

'for many years past,' to quote a letter of Sept. 10, 'been able to report almost every earthquake in all parts of the world, before any news comes in. She suffers from a curious nervous tension as though under the influence of electricity. It almost incapacitates her from work while it lasts, but goes off suddenly. These Japanese earthquakes have affected her badly, and she had a bad attack yesterday. She has consulted several doctors, but none of them has been able to relieve her.'

Almost immediately the explanation of the 'bad attack yesterday' was realised. A letter dated Sept. 11 followed, saying:

'When I wrote yesterday neither my daughter nor I had any outside tidings of an earthquake-but the evening papers reported one in India-precisely at the time when she was suffering most.'

The report was, of course, quite correct: there was a considerable shock in India on Sept. 9 at 11 p.m. of our summer time (22h. 1m. 30s. Greenwich). The testimony is unexceptionable and indicates a line of inquiry which has hitherto not been explored, and may supply unsuspected information both to seismology and to physiology. It would also be applicable to all large earthquakes, and not merely to those which occurred so near a city or inhabited district as to cause damage to buildings and to life. The observations to which Mallet directed attention are practically confined to these latter instances, which have an obviously accidental and somewhat fictitious importance. The almost purely accidental character of the damage done may be illustrated from the Californian earthquake of 1906. It occurred at 5 o'clock in the morning, local time, when the greater part of the city of San Francisco was asleep. In the workmen's quarter, however, the inhabitants had risen and lit their stoves to prepare Vol. 241.-No. 478.

G

early breakfast before going to work. These stoves were overturned and started fires. The fires were not at first serious and might have been extinguished but that the earthquake had burst the water mains so that no hoses could be got to work. Hence the fire slowly and steadily spread, though it was ultimately conquered by somewhat heroic measures. But we see what accidental importance is acquired by the time of occurrence. An hour or two earlier no stoves or fires would have been alight at all, and much damage by fire would have been saved: an hour or two later, all the domestic fires would have been lighted, the conflagration would have been immense, and probably nothing could have saved the whole city.

But nowadays, as already remarked, in addition to the information collected on the spot, which may be affected by what can be fairly called accidental circumstances of this kind, we have available a number of data from seismographs established at stations all over the world. The Californian report quotes records from seventy-six stations, the nearest being the Lick Observatory, only seventy miles away, and the most distant Mauritius on the other side of the globe. They do not include Pulkovo, for the great Russian organisation which did so much for seismology under Prince Galitzin for a few brief years, until its famous leader died under stress of the war, had not yet been started; and most of the instruments were of the pioneer type devised by John Milne, which Galitzin improved almost out of recognition. But with an earthquake of sufficient intensity the most primitive instrument will give an accurate record. The defect of Milne's instruments was that, with a feeble earthquake, they took a little time to get going, depending for their sensitiveness on the principle of resonance'; just as a child giving a swing to another may require several well-timed pushes before the swing begins to move sensibly, whereas a strong man could start it with one vigorous effort.

Galitzin, by beautiful electro-magnetic magnification, flattered the feeble earthquakes so that though they had been before as children, they now seemed like strong men; and he was thus able to abolish the 'resonance' principle; so that his records showed clearly what was happening

at any time, without inheritance from all that had gone before as was inevitable with the Milne machines. He distributed a few of his beautiful instruments over Russia and Siberia, while they found their way also to England, France, Belgium, Holland, and one or two more distant stations. He thus initiated a new accuracy in seismology. Besides this he was expert in the mathematical theory required for the discussion of the records. Altogether a wonderful and delightful man whose loss was an incalculable blow to the science. A meeting of the International Association of Academies at St Petersburg in 1913 gave us the opportunity to see his wonderful installation at Pulkovo, and he had already been to England more than once, especially to the International Congress of Mathematicians at Cambridge in September 1912, when he gave a stirring address (in English-he was master of several languages) on seismology. In May 1915 he accepted nomination as Halley Lecturer in Oxford for the following May, and for many months retained the hope of fulfilling his engagement. But the heavy work of organising the Russian Meteorological War Service broke him down, and he died in August 1916 within a few weeks of the date on which he had hoped to lecture. We can scarcely suppose that he, a member of the ruling house, would have been able to revive Russian seismology. As it was, it died a natural death when existing stores of photographic paper came to an end. They could make no more in Russia and could not afford to buy from abroad. A bright patch among the clouds is that the Petrograd Academy of Sciences managed in 1919 to print Galitzin's last two papers and have sent copies to England. They are of great importance, like all else he did. One, a searching inquiry into the possible causes of microseisms,' will be presently considered. The other is more technical, and concerns the angles at which earthquake waves from a distant station strike up to the earth's surface; but it is of fundamental importance to scientific investigation.

Since this was put in type, a letter (dated Nov. 30) has come from the Russian Académie des Sciences saying that seismological work was recommenced during the latter half of 1923 at Pulkovo, Ekaterinburg, Irkutsk, Platigorsk, Baku, Tashkent, and Cabansk, and in 1924 will be started at Tomsk, Makeyevka, and five other stations in Turkestan and the Caucasus.

The impulse to pause for a few moments to remember so great a loss to the world was well-nigh irresistible; but we must now return to the point at which Galitzin's name drew us aside. He had made a vital improvement in the pioneer instruments which Milne had scattered over the world. To secure the desirable end of getting news from all directions as speedily as possible, Milne had studied simplicity and economy in his seismograph rather than the highest accuracy and efficiency: his motto had been, 'Let us have something which any one can work, even far from civilised appliances.' Stimulated by his enthusiasm cable companies, island governors, and exiles from home set up his machines and sent the records to Milne's home in the Isle of Wight for his collation and discussion. The result was a most valuable preliminary survey of the kind of information likely to be forthcoming. But Milne recognised the great instrumental advances made by his friend Galitzin, and heartily welcomed them. He hoped that his own seismograph might be improved in the same direction, and at the time of his death in 1913 Mr J. J. Shaw of West Bromwich was already at work, with Milne's cordial approval, on these improvements. They were successfully completed a year or two later, and the resulting instruments, under the name 'Milne-Shaw,' are now established in many parts of the world, giving information comparable with that from Galitzin instruments, at a much reduced cost of equipment and working. It would have been quite impossible to obtain the erection of the beautiful but expensive Galitzin instruments at more than a few of these stations, if at any.

We have as yet not given any precise idea of the form of a seismograph, though its action has been compared with that of a swing. It is indeed essentially a swing, though it is more like a swing-gate than a swing which goes with a roundabout. Perhaps the nearest analogy is the horizontal boom attached to a ship's mast, which we have seen swing idly to and fro as the vessel rolls among the billows. It requires a little effort of imagination to picture our solid earth rolling about a seismograph in a similar way; but the analogy is really very close in all but actual scale. An earthquake sends small waves rolling or shivering along the earth's surface,

and also through its interior. The solid seismograph pillar rocks like the boat, and the seismograph boom swings in miniature like that which, on a boat, is apt to injure the unwary. The miniature swings are magnified several hundred times and photographed on a drum which revolves in spiral fashion. The resulting trace usually appears as an even line, though on looking closely one can see the persistent small microseisms.' But there is no mistaking a large earthquake. On developing the film it leaps to the eye,' and I well remember the experience of this kind which startled me on the morning of Sunday, Sept. 2, 1923. For the great earthquake of the preceding day we had been prepared by a telegram from Mr J. J. Shaw, as well as by the news in the morning papers. But Sunday in these postwar days obstructs papers and telegrams, and I was not only quite unprepared by outside news for anything unusual, but predisposed to expect the comparative quiescence which usually follows an exceptional shock. But immediately the developer began to work, there was a great patch on the film, so nearly like that of Saturday that my first thought was of some extraordinary mistake by which the previous day's record had become duplicated. When further scrutiny showed this view to be untenable, and that the second earthquake had really occurred, my mind at once pictured the corresponding bewilderment of Mr J. J. Shaw, whose seismograph rings a bell in his bedroom, at being wakened from sleep at almost the same early hour (4 a.m. summer time) on two nights running. Not till after that did my thoughts travel from the instrumental records to the ghastly possibility that the rescuers at Tokyo and Yokohama might have been themselves overwhelmed in their works of mercy by a catastrophe as appalling as that of the preceding day. For some reason, as yet not quite clear, this was happily not the case. Possibly the slight variation of locality indicated below may have made all the difference; or a greater depth of the shock. But neither of these seems quite adequate. Can it be that further destruction was practically impossible? that houses thrown flat could not be overthrown further? The official report quoted in the 'Times' of Sept. 10 says: