352 Mr. Ivory on M. de Freycinet's Experiments 



significant, the same numbers, about *003 less than according 

 to M. de Freycinet, should express the length of the pendu- 

 lum at Port Jackson. We have likewise a direct determina- 

 tion of the pendulum at this station by the Spanish navigators, 

 whose experiments are calculated by M. Mathieu in the Con. 

 des Terns, 1816*; from which the length comes out equal to 

 39*07682, or '00364 less than M. de Freycinet's experiment.' 

 We must therefore be allowed to demur a little with respect 

 to the accuracy of the pendulum at Port Jackson ; for accord- 

 ing to what has been said, we cannot in this instance safely 

 apply the usual remedy of local attraction. 



At the Isle of France and at Guam and Mowi, the errors 

 appear enormous. But it ought to be observed, that these 

 errors are not the creations of my formula, which, in reality, 

 represents the actual experiments with the same correctness 

 as if the errors had been very small. If the pendulum beating 

 seconds at the Isle of France be transported to Rio Janeiro, 

 in the same hemisphere and nearly on the same parallel, it 

 must be shortened, according to the experiments, '00398 in. 

 in order to oscillate in the same time ; but, according to the 

 difference of latitude, it shoidd be lengthened about *00657 in 

 any hypothesis of ellipticity ; wherefore there is really a local 

 irregularity at the Isle of France compared with other points 

 of the same parallel, affecting the length of the pendulum to 

 the amount of '01080, nearly the same as the error by my 

 formula. And the same thing may be shown with regard to 

 Guam and Mowi, by compai'ing the first with Madras and 

 the other with San Bias. The purpose of a formula strictly 

 deduced from facts without any undue aid from hypothesis is, 

 not to extinguish discrepancies actually existing in Nature, or 

 supposed so to exist, but to exhibit them as they really are. 



On the whole, it does not appear that the experiments of 

 M. de Freycinet bring into notice any good reasons for chang- 

 ing our opinion about the mean figure of the earth, whatever 

 they may do with regard to great irregularity in the distribu- 

 tion of gravity. 



In the Annals of Philosophy for Oct. and Nov., there is 

 an account of an experimental determination of the pendulum 

 at the equator by Captain Goldingham. It must be remarked 

 that the operations were not performed under his immediate 

 inspection ; they were executed by two observers, previously 

 instructed at Madras, and dispatched with written directions, 



* P. 322. The seconds pendulum at Port Jackson being 1*0056969, th; 

 at Paris is found equal to 1-00704998: hence the latter being 39-12929, 

 the former will be 39-07682. 



in 



