my house were visited every day, while those at a distance-as, for example, at Kamakura (C and D) — were visited at least once a week. Notwithstanding the time taken in making journeys to the more distant stations, one of which occupied from ten to twelve hours, because the clockwork kept going for a week and the lamps burned for two or three days, it was often possible to keep six instruments working simultaneously. The notes and photograms for 1893 and 1894 obtained from stations A, FIG. 1.-Map showing Positions of Pendulums. Open ground BM ६ F Pond To SS Smiles To C&D 2 x miles G & H 19 miles, miles Plateau Streets and Platean 60 ft above the plain. Streets and gardens Q. Streets, gardens, open ground. Scale about 0800 Distances in yards from ▲ J-416 R-500 N-1200 Q-3500 E, F, C, D, G, H, and I, which were destroyed by fire, are fortunately described in the fourteenth report to this Association. In order to show the relationship of these observations to those made during the past year at A, H, I, and the remaining eleven stations, they are briefly referred to in the following notes. The sensibilities of the different instruments which are described in the Report for 1894 are indicated by the number of millimetres the end of the boom was deflected by turning one of the screws in the bed plate one degree. The pitch of a screw was one millimetre, and its distance from the axis on which the bed plate was tilted may be taken at 220 millimetres. All the instruments excepting A had usually a sensibility of 1° for 3 mm., which is approximately equal to a tilting of 1" for each millimetre of deflection. Occasionally the sensibility was increased twofold or threefold. The reason for this indefiniteness respecting sensibility is that the notes relating to the calibration of the instruments were burned. The chief value of these determinations is to give angular values for the diurnal wave, and because it will be shown that this is a quantity with large variations depending upon locality and weather, the necessity of accurate measurements becomes more apparent than real. 1°=12 mm. The Pendulum at A.-Sensibility, September 26, 1894. or 1 mm.=0.20. This instrument, which was adjusted to have a sensibility twice or six times greater than the other instruments, was installed in my private observatory on a massive well-built stone column resting upon a bed of concrete. Towards the east and west it was protected by 60 feet of building, but to the north and south by only about 10 feet. The only window in the room, which was on the south side, was always closed by a curtain on the inside and a solid shutter upon the outside. The reason that it was oriented to record N.E. and S.W. motions was because, as was explained in the last Report, there were reasons for believing that in such a direction earthquake and other motion had a maximum. The daily movements were often eclipsed or made indefinite by the occurrence of tremors. When they were visible, although on one occasion they were represented by deviations of 8 mm., it was rarely that they exceeded 2 or 3 mm. In many instances it seems that a second wave was superimposed upon the ordinary diurnal disturbance. Excursions towards the N.E. were usually completed between 5 and 15 hours, while the S.W. motion ended between 0 and 3 hours (fig. 6, p. 134). Pendulums at C and D.-These were installed upon rock in a cave at Kamakura, about 27 miles distant from the pendulums in Tokio. One of these recorded motion in the direction of the dip of the strata, and the other in a direction at right angles to this. Their records are described in the Report for 1893-94. Pendulums at E and F.-These pendulums were in directions N.E. and N.W. or parallel to C and D. The N.W. booms were parallel to A. The installation was in an underground chamber, where as at Kamakura the daily change in temperature was practically unappreciable. These records are described in the Report for 1893–94. Pendulums at G and H.-The pendulums G and H were placed on the rock in a cave at the Yokohama Brewery. Their orientation corresponded to that of C and D or E and F, and their records are referred to in the above-mentioned Report. Pendulum at I.-This pendulum was in an exceedingly damp cave at Kanagawa, about three miles distant from G and H. Its boom, like that of A, pointed towards the north-west. The few records obtained from this station are described with those of the above-mentioned three stations. The Pendulum at J.-Sensibility, August 18, 1894, 1° 6.5 mm., or 1 mm.=0"-39. January 5, 1895, 1°=3 mm., or 1 mm.=0.80. This was an aluminium boom which, with its plate and index, had a total length of about 4 feet 9 inches. Its cast-iron stand rested upon a slab of slate upon the top of a short brick column, which rose from a layer of concrete covering a bed of gravel rammed into the natural earth. It was covered by a coarse wooden case. The whole arrangement was sheltered by a small wooden hut, 9 feet long and 7 feet broad, and up to the eaves 6 feet in height. This hut, like all the other huts, admitted so much light that the photographic films had to be changed at night. Currents of air came in freely, and, as might be expected, there were considerable fluctuations in temperature. On the west side the ground was flat and open, and it was also fairly open towards the north and south. On the east side, however, it was sheltered by a small hill and trees, behind which came a pond, more trees, and then instrument K, which had a small tract of open ground upon its eastern side. The westerly motion, which varied from 5 to 40 mm., usually took place between 18 or 21 hours and 6 or 9 hours, that is to say, the pendulum commenced to move towards the west at about 6 or 9 A.M., and continued this motion until 6 or 9 P.M. (fig. 7). During the night the easterly or return motion was gentle, and usually less than the motion towards the west. On wet cloudy days no curves were visible. Tremors were not marked at this station. Comparing the N.E. and S.W. motions of A with the E. and W. motions at J in 24 instances these movements were completed at about the same hour. In 21 instances, however, there is a difference between them of from 5 to 10 hours. An experiment which was made at this station was to dig a trench round the hut on its south and west sides. This was 5 feet in depth, while its distance from the column was about 10 feet. The only effect that this produced upon the daily diagram seems to have been that the points of inflection in the curve became somewhat sharper, the range of motion of the pendulum remaining constant. Pendulum at K.-Sensibility, September 20, 1894, 1° 4.5 mm., or 1 mm. =0.50. November 21, 1894, 1°3 mm. The installation of this instrument, excepting the fact that it was exposed to open ground towards the east and north, and sheltered by a grove of trees upon the south and west, was similar to that at J. The instrument itself was like that at J. The diurnal movements had a range of from 4 to 40 mm. The westerly excursion usually commenced at 5 or 6 A.M., and continued until about 4 or 6 P.M. (fig. 8). The motion was therefore about one hour in advance of that at J, which roughly corresponds to the difference in time at which the ground in their respective vicinities were exposed to the morning sun. Comparing the hours at which the easterly and westerly motions of J and K were completed in fifteen cases, they closely agree. When these hours do not agree, K has usually reached its western limit from one to four hours before J. In two instances it completed this movement seven hours after J, while in two other cases one pendulum has been near its western limit, while the other has practically completed its movement in an opposite direction. Tremors were not marked. The object in placing J and K, which were 275 yards distant from each other, on opposite sides of a small grove of trees, was with the expectation of finding that at the same time they moved in opposite direction. It is seen that the expectation was not realised. Pendulum at L.-This instrument, which in construction is very like that at A, and is similarly oriented, stands beneath a wooden case on the concrete floor of a cellar in the north-western corner of the Engineering College at the Imperial University. It is without recording apparatus. Its pointer floats over a scale, and its position is noted every day about noon. When first set up on September 19, 1894, it had a period of 28 seconds. Its movements, which are indicated in millimetres, the sign +meaning displacement towards the west and towards the north and east, have been gradual. They were as follows : Pendulum at M.-The object of this pendulum, which was installed upon the same column as A, was somewhat different from that of the others described in this series. It consisted of an aluminium boom loaded at its outer end with a weight of about 300 grms. In addition to this it carried a small vessel of ink from which a capillary tube projected, making the total length of the boom about 4 feet 6 inches. This tube was balanced, so that it barely touched the surface of a band of paper moving at the rate of about 8 inches per hour. The force required to deflect the boom one millimetre when applied at a distance of 4 feet from the agate pivot was approximately one milligramme. Before it was destroyed by fire it recorded the occurrence of several local earthquakes, and, considering its sensibility to tilting, it is probable that it would have recorded the gravitational elastic waves of disturbances originating at great distances. A necessary adjunct to such an apparatus in order to obtain an open diagram is the addition of a quick speed feed for the paper, which must come into action directly the pendulum commences to be deflected to the right and left of its normal position. Such a device was designed for me by my colleague, Mr. C. D. West, and it apparently works more satisfactorily than the original form of this kind of apparatus which is found in the Gray-Milne seismograph. Pendulum at N. Sensibility, January 5, 1895, 1°-3 mm. January 25, 1895, 1°=2.5 mm., or 1 mm.=1".03. The pendulum used at this station was originally at K. The hut was situated on the western side of Tyeno Park, near to its southern extremity. It was sheltered by trees on its eastern and southern sides. On its western side, where it was open, there was a steep scarp leading down to the Shinobadzu Pond, which lies in the bottom of a flat open valley. A, J, and K were on the plateau on the opposite side of this valley, the heights of these stations being about 50 feet above the flat plain on which the greater portion of Tokio is situated. The movements were usually small, seldom exceeding 7 mm. The westerly movement commenced from about 6 or 9 P.M., and continued until about noon next day; that is to say, that about the time when the instruments upon the opposite hill or plateau were going eastwards, the instrument at Uyeno went towards the west, and vice versâ (fig. 10, p. 136). Pendulum at 0.--Sensibility, January 12, 1895, 1°=1.5 mm. January 22, 1805, 1o=1·5 mm. Station O was situated at a place about 20 yards to the south of J. It only differed from the instrument at this latter station in the fact that the boom of the pendulum pointed from west towards the east, and it therefore recorded north and south motion. During the few weeks it was used, it showed a small but regular diurnal fluctuation, being farthest north about 3 or 5 P.M., and farthest south between 9 A.M. and noon (fig. 12, p. 137). Pendulum at P.-Sensibility about the same as N. The instrument used at this station was the same as that used at N and K. The hut faced an open space about 70 yards square on its northern side, but on all other sides it was shaded by high trees. For one or two hours about mid-day a few rays of sunshine reached the roof of the hut and the northern side of the open space. That the ground within 20 or 30 yards of the hut had but little sunshine may be inferred from the fact that after a fall of snow this remained upon the ground for ten or fifteen days. On open ground the snow disappeared in two or three days. During the day this would slightly thaw upon its surface, and at night it would freeze. About 50 yards to the east, a bluff sloped down to the Tokio plain. Observations were only made between January 14 and February 4. The movements were extremely irregular, the most peculiar happening between January 14 and January 26. On these days, excepting the 23rd and 25th, during the night the pendulum made a rapid excursion towards the east, returning to its normal position some time about noon on the following day. These abnormal movements took place upon nights when it was unusually cold, and therefore they may have been due to the freezing of moisture beneath or in the vicinity of the column (fig. 2). Pendulum at Q.--This instrument, which is in charge of the Meteorological Department, is as well installed as the pendulum at A. It stands on a stone column in a dark room in the same building with a selfrecording electrometer. The boom, which at first was partly made of lacquered bamboo, but which has been replaced by one of aluminium, points from north to south. Immediately outside the room there is one of the castle walls sloping down to a deep moat, beyond which comes the plain of Tokio. The diurnal movements are slight, but decided, the westerly excursion being completed at from 3 to 6 P.M., and the easterly at about 6 A.M., which corresponds to the motions at J and K. Tremors are slight. Pendulum at R. This pendulum was set up in a hut in the garden at No. 17, Kaga Yashiki, Tokio. At a distance of about 5 yards on its western side a steep bank leads down to a road, which joins a second road at right angles on the north side of the hut, about 30 feet below the level of the garden. The instrument is intended to act as a seismograph, sensitive to slight vibratory motions, while from the length of its boom, which is about 3 feet, it is also able to record slight changes in level. The first 2 inches of the boom is a small metal tube, at one end of which there is an agate cup. This boom is continued by a reed, at the end of which there is a |