ON THE THERMAL CONDUCTIVITIES OF CEfiTAIN ROCKS. 139 



in the interior of the globe, by instrnmentally measured values of the 

 thermal coefficients to which they either lead, or which, on the other 

 hand, they require for their extensions, are now at least partially sup- 

 plied. But it must be admitted that many imperfections of experimental 

 methods, as well as intricacies of the theory's applications, still remain to 

 be removed before the confirmation and assistance which one branch of 

 the subject of heat-conduction can derive from and may usefully offer to 

 the other, can yet be satisfactorily affirmed to be complete. 



Upon this point, which embraces the geological aspect and objects of 

 the present investigation in its widest sense, much more might be written 

 and comprised in this report which the Committee has noted and col- 

 lected in its work of original reference and research. But as these notes 

 and historical reviews would lead to the extension of this report much 

 beyond its contemplated length, the Committee deems it necessary, in its 

 concluding report, to confine itself to the results of recent work and pub- 

 lications which bear most immediately upon the experimental part of 

 the problem regarding whose progress and recent investigations it has, 

 during the long period now terminated of its frequent I'eappointment, 

 chiefly endeavoured to record the nseful acquisitions, and to present to- 

 gether an outline of the best existing information. 



APPENDIX. 



Abstracts of Papers relating to the Conduction of Heat, and a List of 

 Authors' Names and Writings on the Subject. By J. T. Dunn. 



Dalton, • Ann. Chim.' (1), 45, 177 (1803). — Method of determining point of maxi- 

 mvun density of HoO by thermometers. 



Despretz, 'Ann. Chim. Phys.' (2), 19, 97 (1821).— The first of the well-known 

 memoirs. Eesults : Cu 12 ; Fe, Zn, Sn, 6 ; Pb, 2 ; marble, ^ ; brick clay, 

 porcelain, 35- 



Despretz, ib. (2) 36, 422 (1827). — Extension of former results, with numerical details 

 of experiments. Good conductors satisfy Fourier's exponential law, but the 

 successive quotients with bad conductors differ widely. Figures : — 



Porcelain 12 -2 



Brick and 'i -i -i . 



firebrick/ ^ ^ 



Fourier, ib. (2) 37, 291 (1828). — Conducting powers of thin films. A thermometer is 

 immersed in Hg in a conical iron vessel with a flexible skin-bottom. This 

 having been warmed to a suitable temperature is placed on the film or lamina 

 lying on a block of constant temperature, and the fall of the thermometer noted 

 Comparative results only. An air-thermometer has the lower half of its bulb in 

 contact with the lamina, which again is in contact with a hot body of constant 

 temperature, while the upper half of the thermometer-bulb is in contact with 

 melting ice. Order of conductivity given by the temperature which the ther- 

 mometer takes up. 



De la Bive and Decandolle, ib. (2) 40, 91 (1829).— Conductivity of woods. Same 

 method as Despretz. End of bar cased in tinplate and heated by spirit lamp. 

 Order of conductivity only. Allier ( Cratcegus aria), walnut, oak, pine, poplar 

 (long.). Walnut, oak, pine (trans.). Cork (long.). 



Lam§, ib. (2) 53, 190 (1833). — Theoretical remarks on isothermal surfaces. 



Despretz, ib. (2) 71, 206 (1839). — Liquids. Heated from surface in deep cylinder. 

 Temperatures of thermometers follow same law as in solid bar. At same level 

 temperature diminishes from axis towards side, and thence to middle of wall. 

 With cylinders of different diameters temperature-differences agree with theory. 



