THE LAWS OF CONDUCTION OF HEAT IN BARS. 75 



I hope rarely so. Much larger corrections would have been inevitable at the 

 highest temperatures (about 200° Cent.), had I not invariably employed for these a 

 thermometer in which about 110° of the mercury was expelled from the bulb into 

 the cavity at the top of the stem. The corrections for the reading of this thermo- 

 meter were determined by Mr Welsh with extraordinary care. As its indications 

 only commenced about the boiling-point of water, the length of the column 

 exposed to the air was comparatively short. 



45. For the principles on which the experimental investigations are founded, 

 I refer to Art. 5, &c, of the former part of this paper. It will be recollected that 

 there are two distinct classes of experiments, in one of which (the statical) the 

 permanent temperatures at different points of a bar are to be observed ; in the other 

 (the dynamical) the velocity of cooling of a short bar of similar section, uniformly 

 heated at first, is to be ascertained. I shall now proceed to describe these experi- 

 ments severally more in detail than I have yet done, and to classify and discuss 

 the results. 



§ I. Statical Experiments. 



46. The Apparatus.— A general account of this has been given in Arts. 

 1.7-20. It will, however, be rendered more intelligible by a reference to Plate I., 

 fig. 1. The long wrought-iron bar AB was supported on a wooden frame CD 

 by means of one fixed support E, and two moveable props F, Gr, which were all 

 of wood, and were brought to a blunt edge at top, on which the bar rested, at 

 about 15 inches above the top of a massive table, which stood in a spacious 

 apartment attached to the Natural Philosophy Class-room (Edinburgh Uni- 

 versity). No fire was allowed during the experiments, and the south shutters 

 being closed, the room was lighted from the north. At the end of the 

 bar, towards the left side of the figure, was attached the heating apparatus, 

 a cast-iron crucible H, usually filled with just-melting lead. It was kept 

 hot by means of the powerful gas-furnace I, with a double metal chimney 

 and two concentric rows of burners. The gas was derived from the main pipe 

 by a flexible tube L, and passed through one of Milne's patent gas regulators, K, 

 with a view to obtaining a uniform flame, which, however, remained subject to 

 occasional fluctuation. The connection of the crucible with the conduction bar 

 will be best understood from the sectional diagram in Plate II. fig. 1. An 

 internal flange a, a' was cast on the crucible, leaving a square cavity 2-5 inches long, 

 into which the extremity A of the conduction bar was thrust, and was retained 

 there by friction only. The exterior face of the crucible b c is almost vertical, 

 and determines the position from which the distances of the thermometers along 

 the bar are reckoned. Supposing the crucible itself to be maintained at the constant 

 temperature of melting lead, it seems reasonable to assume that the bar A, so far as 

 encased within it — that is, up to the zero line b c — may be regarded as having nearly 



