89 



III. CONCLUSIONS.* 



(a) CHANGES IN THE VERTICAL. 



CHANGES in the position of the outer end of the pendulum, which is an aluminium boom three feet in 

 length, have been measured on the seismographic films at intervals of four hours, and in certain instances 

 every 30 minutes. These films are strips of bromide paper each 2 inches in width and 35 feet in length. 

 They moved beneath the end of the boom at a rate of 60 millims. per hour. The total length of film 

 brought home by Mr. BERNACCHI is about 3000 feet. One millimetre deflection of the photographic trace 

 of the outer end of the boom is approximately equivalent to a tilt of 5". 



The measurement of the displacement of these traces was undertaken by my assistants, Mr. SHINOBU 

 HIROTA, and Mr. HOWARD BURGESS, of Newport, and it is in consequence of their assistance that the 

 analyses of these records have reached their present stage. The results are at present in two forms as a 

 manuscript register and as a series of curves drawn on squared paper. They are in charge of the Royal 

 Society. Before the analyses of these can be completed they must be supplemented with corresponding 

 records from barographs and thermographs. The times of total darkness, continuous light, sunrise and 

 sunset have already been entered on the squared paper. Also, as Mr. BERNACCHI remarks, tidal 

 fluctuations, ice movements, changes in volcanic activity may also hold some relation to the wanderings of 

 the pendulum. It is, therefore, desirable that information relating to these phenomena should be obtained. 



An examination of the curves indicates that there have been many comparatively large and rapid 

 deflections of the pendulum, particularly after its removal from the magnetic observatory to the living 

 hut. For example, subsequent to the removal, tiltings of 10" have taken place in 20 hours. Displacements 

 of this magnitude suggest a yielding of the foundations or parts of the brick column on which the 

 instrument was installed. My own experience is that in England it takes about 12 months for a masonry 

 pier to become stable. A pier made with a glazed earthenware drain-pipe has only its foundation to settle, 

 and becomes stable more quickly. 



There are other deviations which may be seasonal, whilst others have accompanied marked barometric 

 fluctuations. At certain periods there have also been changes in position of the boom, indicating tilts of 

 5" to 1 0" which have approximately a diurnal periodicity. 



In " Discovery " local time the western excursion of the pendulum was most frequently completed 

 about 11 p.m., whilst it was usually farthest east about 3 p.m., and this took place whether there was sun 

 or no sun. To explain these changes, possible distortions produced by sun heat on the earth's surface have 

 been suggested. 



That an accumulation of a water load in a valley apparently causes its two sides to approach each other, 

 whilst a body of men approaching an observatory will cause a pendulum inside the same to swing towards 

 the advancing load, have strengthened the suggestions that changes of level observed at a station might be 

 influenced by differences in evaporation or of vegetable transpiration on opposite sides of such a building. 

 These suggestions, although they do not directly bear upon work carried out in the Antarctic regions, 

 have received attention.! 



Another suggestion which I venture to make, and it is one which, for many reasons, I think deserves 



* This section is reprinted, with alterations, from ' Proceedings of the Royal Society,' series A, Tol. 76, 1905. 

 t See ' British Association Reports,' 1895, pp. 115-139, and 1896, pp. 212-218. 



N 



