Natural- Pkibsophical Collections. 463 



or to that of the air, which diminishes very little from the equator to the 200° of 

 latitude; that of the earth, constantly decreasing, must be inferior. 



3d. The decrease of the temperature of the earthj according to the latitude, is 

 represented by the equation. 



a — b sin. 2 I = t. 

 or a and 5 are unknown constants which are to be determined; / the latitude, and 

 / the temperature of the earth. 



M. Kuppfer does not admit, with M. de Buch, that the temperature of the soil 

 can be explained by the absorjition of pluvial waters. 



M. K. remarks, that although the formula represents perfectly the observations ; 

 like all formula with arbitrary constants, whose value is determined by observa- 

 tions, it can only give approximations, and will be inexact on points far from 

 those of the observations. If the minimum were not placed at the pole, but a lit- 

 tle on this side, the formula would not give it, since it is at its smallest value 

 when t =r 90 ; it would only represent the observations to the minimum of tem- 

 perature, and would give too low a temperature to the pole. 



It results from the tables given by M. Kuppfer, that the temperature of the 

 earth at the equator is greater in the middle of Africa than in the sea between 

 Afrka, the East Indies, and New Holland : that it is less on the western shores 

 of Africa and the western coasts of America ; and that in the centre of America 

 U has a considerable value. Under the meridian of Africa the isogeothermal 

 lines to 50° lat. present a marked convexity, whose summit is directed towards 

 the north. Under this meridian are the two active volcanos, ^tna and Vesuvi- 

 us, many hot springs, and almost every where basaltic and porphyritic rocks. 

 Perhaps the maximum of the temperature of the earth is not at the equator ; but 

 this hypothesis is founded only on one observation in the southern hemisphere at 

 Congo. — See Mr. Kuppfer^ s paper at length, Ann. de Chitn. et Phys. Dec. 1829. 

 New Preparation of Caoutchouc for various purposes. 



To THE EdITOU, &C. 



Will you do me the favour to mention, in the next number of your 

 Journal, that I am about to lay before the Society of Arts a new method of 

 preparing caoutchouc, so as to render it useful in various ways, and more espe- 

 cially suitable for covering preparation bottles and jars. It may be easily follow- 

 ed by any one, and is cheaper in the end than any other substance. The 

 elastic covers cling firmly to the body on which they are extended, may be re- 

 moved and re-applied at pleasure, and completely prevent the evaporation of al- 

 cohol, water, and of those fluids which have no chemical action on caoutchouc. 



I shall soon make public, through the medium of your Journal, an adaptation 

 of my method to the putting up of preparations in turpentine, and other fluids 

 whicli dissolve caoutchouc when in contact with it, which, by preventing evapora- 

 tion, will, I trust, prove valuable to the collectors of anatomical preparations. I 

 remain, &c. MoNTGo:MEaY Robertson. 



22, Great King St. Edinburgh, July 28, 1830. 



Observations made relative to the Figure of the Earth, during the Voyage of 

 the Coquille. M. Duperrey was provided with two invariable copper pendulums, 

 •which had been previously employed in the voyage of the Uranie. They were 

 tried at Paris before the departure, and after the return of the expedition ; at 

 Toulon, whilst the vessel was preparing ; at the Malouin Isles, in 54° SP 43" of 

 S. lat. ; at Port Jackson, on the east coast of New Holland ; at the Mauritius ; 

 and at the Isle of Ascension, within the tropics. M. Matthieu has deduced 

 from his calculation of the observations at the Malouin Isles and at Paris, 

 this important result, in opposition to the old opinion, that the north and south 

 terrestrial hemispheres have very nearly the same form. From ]\I. P'reycinet's 

 observations it results that there is, at the Mauritius, a cause of local attraction so 

 intense that it altered the time of a chronometer 13 or 14 seconds daily. It is 

 obvious how interesting it would be to observe, in M. Duperrey's experiments, 

 whether the accidental iixfluence has been as manifest. Fourier, Rapport. 

 VOL. II. 3 P 



