132 MM. Aug. de la Rive and Alph. De Candolle on the 



cury, into which a thermometer was immersed. One of the ex- 

 tremities of the piece of wood was sunk in a tin case, about 2^ 

 centimetres long, so as not to cover any of the holes. This 

 apparatus was freely suspended in the air, and a spirit of wine 

 lamp was placed under the extremity, armed with tin. The 

 flame could only strike this part, on account of the chimne_y of 

 the lamp and plates of glass which we placed vertically between 

 it and the piece of wood, taking care to renew them as soon as 

 the heat began to traverse them. In this manner, the source 

 of heat was single, without, however, directly striking the wood 

 in such a manner as to burn it. In order that the thermometer 

 should have precisely the temperature of the interior of the 

 wood, we threw a little lycopodium powder upon the orifices of 

 the holes, which prevented all external radiation of the balls of 

 the thermometers, and of the mercury which surrounded them. 



At the end of from one to two hours, each thermometer had 

 attained the maximum of temperature which its distance from 

 the source of heat, and the conduclibility of the wood, com- 

 bined with the radiation, permitted it to assume. We only con- 

 sidered the experiment as ended, when the thermometers had 

 attained their fixed point for ten minutes or a quarter of an 

 hour. We have retrenched from all the thermometrical heights 

 the temperature of the ambient air, which only varied from- 6° 

 to 10° centigr. 



The kinds of wood which we tried are six in number, of 

 which three were in the two directions of the fibres. Placed 

 in the order of their conductibility^ beginning with the best 

 conductors, they are the following : — hornbeam, chesnut, oak, 

 fir, poplar, all in the direction of the woody fibres ; then ches- 

 nut, oak and fir, in the contrary direction ; and, lastly, cork. 



On comparing the two extremes, it is found that, in the horn- 

 beam, a very hard and heavy wood, the first thermometer being 

 at 83°, the second was at 4?5°, a little more than the half ; while 

 in the cork, the first being at 78°, the second was only at 14°, 

 a little more than the fifth. The densest woods were in general 

 the best conductors. Chesnut, however, is a somewhat better 

 conductor than oak, although it is lighter. It is also seen from 

 the table which follows, that there is little difference between 

 the woods cut in the same direction, and that their slight homo- 



