as deduced from Experiments with the Pendulum, 1 73 



tleman, the observed pendulum at Spitzbergen falls short of 

 the calculated one 3 \ vibrations in a day ; — in my table the de- 

 fect is only 2{. The density of the matter near the surface of 

 the earth, varying at the different stations of the pendulum, 

 must in some degree influence the time of vibration : but the 

 speculation is premature ; and there is good reason to think, 

 at least in far the greater number of cases, that the effect in 

 question will never be separated from the unavoidable errors of 

 observation. 



What has just been said will not, however, apply to the five 

 anomalous experiments, all of which show an excess of gravity 

 far surpassing the ordinary measure of experimental error. 

 Besides, at two of the stations the pendulums have been veri- 

 fied by two independent experiments varying little in the re- 

 sults. It is remarkable that in all the five cases, and at Gala- 

 pagos, there is an excess of gravity; and, according to my 

 table, there is a like excess, though much less in quantity, at 

 Rawak, Sierra Leone, and Jamaica. On the other hand, there 

 appears to be a defect of gravity at Maranham, Trinidad, and 

 San Bias. Thus there are nine instances of excess, and only 

 three of defect ; and we may suspect that there is some cause 

 of error, different from local attraction, tending to make these 

 tropical pendulums longer than they should be. At the Isle 

 of France, the excess of gravity is no less than 10 vibrations 

 a day; yet this is a small island, surrounded on all sides by an 

 extensive sea; and we might infer, a priori, that a pendulum 

 placed upon it very little above the level of the sea, instead of 

 being accelerated, would be retarded by the great defect of 

 density in the waters of the ocean. But the purpose of this 

 paper is to deduce from the experiments the consequences that 

 necessarily flow from them by a strict investigation ; and I 

 shall refrain from entering into the region of conjecture and 

 opinion. 



Feb. 11, 1828. J. Ivory. 



XXVIII. The Climate of Penzance, Cornwall. — Meteorologi- 

 cal Residts of the Temperature, Wind and Weather deduced 

 from diurnal Observations made at Penzance for 21 Years: 

 {the Thermometrical Observations made at 8 a.m. and 2 p.m.) 

 to which are added the Maxima, Minima, and Media of the 

 Register Thermometer for 7 Years, with the Inches of Rain 

 fallen during that Period. By Edward Collins Giddy, 

 Curator of the Cabinet of the Royal Geological Society of 

 Cornwall. 



January 



