454 E. D. Preston — Study of the Earttis 



that besides the influences just mentioned some allowance 

 should be made for the matter lying above the sea level ; and 

 his formula, based on the relation between the mean density of 

 the earth and that of the crust, is still employed. The pro- 

 priety of this method of treatment has, howevei, several times 

 been questioned. In the first place, because observation seems 

 to show mountains and table lands to be much too light, and 

 in the next place, because the excess of gravity noticed at 

 island stations is very nearly accounted for by making a cor- 

 rection for the downward attraction of the sea, Of course we 

 meet with many anomalies in gravity determinations, but it 

 would seem better in the present stage of the subject to make 

 some disposal of the influence of the continents. The estimate 

 may indeed be a rough one, afterwards to be modified as more 

 data accumulates, but in the light of our present knowledge we 

 may apply the corrections as follows : at a station say 1900 feet 

 above sea level a seconds pendulum will lose eight seconds 

 daily on account of its elevation, and it will be accelerated in 

 the same time by three seconds from the influence of the 

 mountain matter, so that both effects together would cause a 

 loss of five seconds per day. This rule of course does not hold 

 strictly at great elevations ; nevertheless it was found approxi- 

 mately true in the Sandwich Islands where the pendulum was 

 carried to an elevation of about 10,000 feet. Here we had a 

 daily loss of 41 s from elevation and an acceleration of 13 s from 

 the mountain attraction giving a total loss of 28 s daily. This 

 agrees tolerably with the ratio above stated. 



Changes of Latitude, 



Closely connected with the variations of the force of gravity 

 are the changes of terrestrial latitudes. Whether the earth's 

 crust is floating on the plastic or semi-plastic nucleus, and 

 really shifts its position with reference to the axis of rotation, 

 or whether latitudes change by reason of the moving of quan- 

 tities of water and air, it is now impossible to say. Within 

 the last year it has been abundantly demonstrated that latitudes 

 may have an annual variation of a considerable fraction of a 

 second. The International Geodetic Association of Europe 

 has taken the matter up and will send an observer to Honolulu 

 in order that simultaneous observations may be made on oppo- 

 site sides of the earth. The U. S. Coast and Geodetic Survey 

 has been asked to cooperate and will also send an officer to the 

 Sandwich Islands, besides observing continuously at Wash- 

 ington, San Francisco and other points. The Royal Obser- 

 vatory at the Cape of Good Hope will engage in the work and 

 probably other permanent stations may be established in the 

 southern hemisphere. 



