.182 



NATURE 



[December 26, 1895 



and evaporation, its amount varies daily, and with locality, and 

 on two sides of a given station it reaches a maximum at different 

 times. 



Experitnents bearing upon these Suggestions. 



(1) Ten men and boys, representing a load of about looo lbs., 

 at a distance of 1 5 feet from the stone column carrying the pen- 

 dulum in my house, causes it to move as if the ground had been 

 depressed upon the side of the load. 



(2) By quickly emptying a well, which is distant about 100 feet 

 from the above column, of about two tons of water, which was 

 run off down a hill, the pendulum moved away from the well, 

 behaving as if the ground had been relieved of a load on that 

 side. 



(3) A self-recording tide-gauge was set up in an unused well 

 80 yards distant from the underground chamber. This showed 

 that the water in the well rose and fell i to J inch twice every 

 24 hours. The times of sinking, which occur in the morning 

 and the evening, may correspond with the times at which the 

 most water is being drawn from wells in the city. I have not 

 determined whether these movements have any influence upon 

 a small wave which is often superimposed upon the large 

 diurnal wave. 



(4) I find from experiment that on fine days the ground in my 

 garden may lose by evaporation 4 or 5 lbs. of moisture per 

 square yard, or from an area measuring 20 by 20 yards about 

 I ton. 



I also observe that during a bright fine day that, in most cases, 

 pendulums move away from areas that are being relieved of 

 loads in this manner, and which may therefore be rising. 



(5) Because it often happens that a board which has been 

 lying on the grass all night is found in the morning to be wet on 

 its under side, whilst all around the grass may be dry, the 

 following experiment was made on sub-surface condensation. 

 Two shallow trays about i foot 6 inches square were filled with 

 earth, and placed on a flat surface of earth in my garden. One 

 of these trays had a bottom of sheet tin, whilst the other had a 

 bottom of fine wire netting. These were weighed morning and 

 evening, and it was found that sometimes the box with the fine 

 wire netting had increased in weight, while the other had not 

 changed. The inference from this is, that during a hot day the 

 ground is sensibly heated to a depth of one foot. This was 

 proved by diagrams from self-recording thermometers, which for 

 periods of a week had been buried at varying depths. After 

 sunset the surface of the ground is quickly chilled, and aqueous 

 vapour rising into this is condensed to augment the surface 

 weight. 



An open area, which during a day may have lost more weight 

 by evaporation than a neighbouring area which is covered, will, 

 if both areas are connected with the same subterranean water 

 supply, gain the most in weight, not only by an action of this 

 sort, but also by the condensation of moisture as it escapes from 

 the surface, and by the precipitation of moisture from the atmo- 

 sphere. These actions may in part account for the retrograde 

 motion of pendulums during the night. Although differential 

 evaporations and condensations are apparently sufficient to 

 account for certain observations made in Japan, other observa- 

 tions exist to which they are not so apparently applicable. The 

 cause of diurnal waves is therefore not yet known with certainty. 



(3) Earth Tremors. 



The continuous photographic records taken in Japan have 

 thrown much new light upon the occurrence of tremors, while 

 an experiment made at Shide in the Isle of Wight, where a light 

 horizontal pendulum has been established to record earthquakes 

 having their origins in distant localities, has indicated a cause 

 of movements which probably have often been attributed to 

 movements of the ground. 



We know that tremors are more frequent during the winter 

 than during the summer, that they are frequent with a low 

 barometer, and still more frequent when the locality of observa- 

 tion is crossed by a steep barometrical gradient. From these 

 latter accompanying conditions, it may be inferred that tremors 

 may occur whenever a strong wind is blowing, although near to 

 the observing station the atmosphere may be perfectly calm. 

 Recent observations in Japan have shown that tremors are more 

 frequent during the night than they are during the day, and that 

 maxima of motion are reached between 5 and 9 a.m. In fact, 

 some instruments have always shown movements about these 

 hours. One very important observation is that the greatest 

 movements have been recorded by the lightest pendulums. For 



NO. 1365, VOL. 53] 



example, a pendulum having a boom from one quarter to four 

 inches in length is a better tremor recorder than one which has a 

 boom several feet in length. 



The pendulum at Shide was set up upon a newly-built brick 

 pier. Because moisture was given off from this, the inside of 

 the covering case became exceedingly damp, and the bromide 

 film became sticky. To overcome this difficulty, two trays of 

 calcium chloride were placed inside the covering. A few 

 minutes after this it was observed that the pendulum commenced 

 to swing. When the calcium chloride was removed the pen- 

 dulum came to rest. This experiment was repeated several 

 times, and the conclusion arrived at was that rapid desiccation 

 produced air currents of sufficient intensity to cause a light pen- 

 dulum to move, and the diagrams of these movements are not 

 distinguishable from those attributed to earth tremors. Although 

 experiment has shown that differences in temperature of the 

 walls enclosing a light pendulum will cause the same to swing, I 

 have no reason to think that movements due to such causes have 

 been recorded. All that can be said at present is that a differ- 

 ence in the rate at which moisture is absorbed by or evaporated 

 from the different walls of a casing covering a light pendulum, 

 may cause the same to swing. Before we should attempt to 

 explain why such movements are marked at particular hours, 

 and occur with certain meteorological conditions, the necessity 

 of further experiments is obvious. Air currents can hardly 

 explain a set of tremors lasting several hours, where a seventeen 

 seconds pendulum moves back and forth with uniform amplitude 

 and a uniform period of two or three minutes. 



(4) Earthquakes. 



At Kamakura, on the hard rock, the greatest earthquake motion 

 has been given by the pendulum which records tilting parallel to 

 the dip — suggesting the idea that in this direction there is an easier 

 yielding (like the opening and shutting of a concertina) than 

 there is in a direction parallel to the strike. The movements, 

 even for unfelt earthquakes, are sometimes as much as 40 mm., 

 and a disturbance may continue for several hours. On March 22, 

 I and my colleague, Mr. C. D. West, watched an earthquake 

 for I hour and 47 minutes, during which time the pendulum 

 did not swing, hut wi&s forced backwards and forwards intermitp 

 tently, and with extreme irregularity. These earthquakes are ir» 

 the form of earth waves, and usually come from a great distance. 

 A sharp shock which may be felt throughout Tokio and at many 

 places in the country, does not disturb the pendulums, and it is 

 difficult to find a blurr on the photographic trace. 



As in previous years, before certain local earthquakes I have 

 observed abnormal tilting, an explanation for which is suggested 

 in the section on the wandering of pendulums. It is seldom 

 that abnormal tilting has taken place without local disturbances. 

 Local disturbances, unless they are large, which fortunately is of 

 rare occurrence, are only recorded by seismographs. 



One disturbance which I recorded at three stations in Japan 

 was one which had its origin near to the antipodes of that 

 country in the Argentine Republic. The conclusion to be de- 

 rived from this and other observations is that a large earthquake 

 may be recorded at any point upon the surface of the earth. The 

 preliminary tremors seem to reach distant places with a velocity 

 twice or three times that with which mechanical vibrations can 

 be transmitted through glass or steel. Possibly they come 

 through our earth, which therefore may have a higher effective 

 rigidity than hitherto supposed. The undulatory motions which 

 follow the tremors may be transmitted as surface quasi-elastic 

 gravitational waves. Their velocity of propagation lies within 

 the limits of expectation. The chief object of the instrument 

 established in the Isle of Wight is to determine whether it is 

 sufficient to record the unfelt movements due to earthquakes 

 originating in distant localities. 



The first certain records of earthquakes having their origin at 

 great distances were obtained by Dr. E. von Rebeur-Paschwitz, 

 whose recent death has deprived the world of one of the most 

 active workers in the new field which has been opened to 

 seismologists. John Milne. 



THE ANTWERP METEOROLOGICAL 

 CONGRESS. 

 "nPAKING advantage of the attendance of numerous visitors, 

 -*■ scientific and otherwise, that the Antwerp Exhibition was 

 likely to attract, it was proposed to hold, under the auspices of 

 the Geographical Society of that city, a congress on meteor» 



