10 E. D. Preston — Measurement of the Peruvian Arc. 



the total length of the arc more than a fraction of a second at 

 each end. Therefore the question of azimuth is not one of 

 vital importance. 



We now come to perhaps the most interesting part of Bou- 

 guer's work. Not satisfied with investigating the exterior 

 shape of the earth, he determined to study also its interior 

 condition. It had been known for more than fifty years, that a 

 pendulum oscillated more slowly at the equator than near the 

 pole, and finding himself not only in a latitude where the 

 force of gravity was the least, but also in a country where 

 there were exceptional facilities for the study of this force at 

 great elevations, he deemed it his duty to devote some time to 

 the investigation of the subject. Of the two methods, either 

 comparing the times of oscillation by the same pendulum, or 

 comparing the lengths of two different pendulums, vibrating 

 in the same time, he chose the former. In this he has been 

 followed by all later observers. Of course his results cannot 

 now be regarded as of very great value, both on account of 

 the unsuitable methods and inferior instruments employed. 

 But the work pointed in a certain direction, which has been 

 confirmed in a general way by some more recent and accurate 

 determinations. His method of getting time would not now 

 be used in gravity observations, although modern instruments 

 would increase the accuracy of the result several fold. It is 

 doubtful whether, with his instruments, and often using single 

 altitudes of stars, the time was correct within several seconds. 

 About one and one-half hours may be taken as the duration of 

 a swing, so that with the uncertain clock correction, and the 

 short duration of the experiment, great discrepancies in the 

 individual swings were unavoidable The pendulum was of an 

 inaccurate type and its length was found by simply holding an 

 iron bar by the side of a thread stretched by an ounce weight. 

 Contact with the clamp above and the bob below was examined 

 either by means of a magnifying glass or by the naked eye. 



The individual results, for the length of the seconds pendu- 

 lum at Quito, where the conditions were favorable, have a 

 range of about 1/6000 part. This would correspond to dis- 

 crepancies in the times of one oscillation of less than one unit 

 in the fourth place of decimals. Under the circumstances the 

 accordance is good. 



The result was corrected for buoyancy and temperature. 

 The former was here applied for the first time in pendulum 

 observations. It was estimated that the density of the air on 

 the top of Pichincha, was one eleventh thousandth of that of 

 the metal composing the pendulum bob, and since gravity 

 varies inversely as the length of the seconds pendulum, the 

 length found was increased in this same ratio. No correction 

 seems to have been made for the amplitude of oscillation. 



