100 REPORTS ON THE STATE OF SCIENCE. 



instrument has been installed. They may, however, be due to the general 

 warping of a district under the influence of solar radiation, or to the 

 differential effects of loading and unloading of portions of the same. 

 During the day the sides of a valley covered with vegetation lose load 

 by evaporation and transpiration, and therefore underground drainage, 

 tending to carry a water load to the bottom of a valley, is reduced. At 

 night, with the cessation of these processes, the load at the bottom of a 

 valley is increased. At that time streams and certain wells carry their 

 greatest quantity of water. Tt is therefore at night tliat a valley may be 

 expected to sag downwards, a suggestion that finds support in the observa- 

 tion that during wet weather, when we see streams in flood, the sides of 

 the bounding valley approach each other in a marked manner. 



The conclusion is that as the world turns before the sun its surface 

 is measurably smoothed, whilst at night the frecklings on its face are 

 measurably increased. 



VIII. Antarctic Earthquakes. 



From March 14, 1902, to December 31, 1903, a horizontal pendulum 

 of the British Association type was installed in Victoria Land, about 

 fifteen miles distant from Mounts Erebus and Terror. The instrument 

 was in charge of Mr. Louis Bernacchi, who was attached to the 

 s.s. ' Discovery.' Although observations were made under exceptional diffi- 

 culties, Mr. Bernacchi brought back about 3,000 feet of photographic film. 

 This was examined by your Secretary, Mr. Shinobu Hirota, and Mr. 

 Howard Bui-gess of Newport. ' Preliminary Notes ' on this analysis are 

 to be found in the 'Proceedings of the Royal Society,' vol. A76, 1905, 

 pp. 284-295. The more important results are as follows : — 



1. Out of 136 records of earthquakes seventy-three refer to disturb- 

 ances Avhich originated in a sub-oceanic region between New Zealand and 

 Victoria Land. In the British Association maps of Seismic Distribution 

 this TiQ'fr district is indicated by the letter M. The greatest frequency 

 was in the months April, May and June. 



2. Certain earthquakes from district M have been recorded to the 

 south-east by the ' Discovery,' and along a band about 20° in width 

 towards the north-west as far as Britain. This phenomenon of recordable 

 motion being propagated in one direction only round the world is now 

 known to be true for earthquakes originating in other districts. 



3. An earthquake may be recorded at stations near its origin and its 

 antipodes, but not at intermediate stations. 



4. As an earthquake radiates, the phase of motion which travels the 

 greatest distance is that of the largest waves or P3. 



5. The average arcual velocity for P3 is approximately 3 Km per 

 second, with a possible acceleration in the quadrantal region of its path, 

 where it may reach 4 Km per second. 



6 . Other results refer to slow changes in the vertical, diurnal waves, 

 tremors and pulsations. 



IX. Note on the Determination of the Time of Origin of Earthquakes. 



By R. D. Oldham. 



One of the greatest desiderata in the study of the nature and 

 propagation of earthquake waves is an exact determination of the 

 time of origin of the earthquake. Few are the instances where a great 



