±±3 K D. Preston— Study of the Earth's 



of an error in the time of an oscillation is doubled when it is 

 referred to the length of the pendulum ; so that the weakest 

 part of the entire investigation is the length of the time of one 

 oscillation. For this reason it has been assumed that the pen- 

 dulum can never be made to compete with metal bars in giving 

 us a uniform standard of length. Whether it is better to refer 

 our standards to the wave-length of light, or to some material 

 thing involving permanence of capacity, is not in the exact 

 line of our thought at present. Objections have been urged 

 against both these methods. The former because it was sup- 

 posed the earth might eventually move to a region of space 

 where the wave-length of light was different, and the latter 

 because a capacity-measure may change its three dimensions 

 unequally so that the permanence of its capacity would be no 

 proof of the permanence of one of its linear dimensions. 

 What most concerns us is to know that the seconds pendulum 

 is not our best standard of length. 



Method of Study. ' 



The most advantageous way of treating the figure of the 

 earth is now conceived to be different from that employed 

 hitherto. The measurement of an arc of the meridian by 

 triangulation is not the best means of arriving at the flattening 

 although it gives the actual size of our planet with an accuracy 

 fully equal to the requirements of the case. The problem 

 should be separated into two distinct investigations, and to 

 each one of them methods should be applied, that will deter- 

 mine the unknown quantities to the best advantage. Let pen- 

 dulum observations and the moon's parallax determine the 

 flattening and let triangulation measure the actual size. Then 

 we shall have each method working in a field where it has the 

 greater power and the results will be correspondingly better. 

 When the flattening is determined its value may be used as a 

 known quantity in the equations that determine the length of 

 the axes. This method is suggested by Professor Harkness. 

 Admitting then that pendulum observations must be employed 

 for the study of the earth's figure, what is the best way of 

 doing it ? Here we are confronted by two distinct schools. 

 The Germans, followed by the Russians and Swiss, have 

 always favored absolute determinations of gravity. The Eng- 

 lish stand as the exponent of the rival school and only meas- 

 ure differentially. The U. S. Coast and Geodetic Survey has 

 followed both methods to some extent, and now seems to be 

 favoring differential measures especially for a preliminary sur- 

 vey of the country. The difference between the methods is 

 briefly this : the Germans measure the actual force of gravity 



