34 Dr. T3mdall on the Progress of the Physical Sciences : 



and Franz the same object is attained by silvering the bars. As 

 the latter were thin, a very necessary precaution was to shelter 

 them from currents of air, and to preserve the surrounding air 

 at a constant temperature. The arrangement also permitted of 

 the bars being examined in vacuo. The following description 

 will, we doubt not, render the manner of experiment sufficiently 

 plain without having recourse to a figure. 



Each bar was caused to occupy the axis of a cylindrical vessel 

 of glass, being retained in this position by causing the ends of 

 the bar to rest in two tubes fixed in the ends of the cylinder. 

 The tube at one end was extended, and surrounded by a space 

 which commiinicated with a vessel of boiling water. When this 

 space was filled with steam, heat was communicated to the tube 

 and to the bar within it, whence it propagated itself through 

 the mass of the bar. Through the opposite end of the glass 

 cylinder, and at some distance from its centre, a rod was caused 

 to pass which was capable of sliding up and down air-tight. To 

 the end of this rod was attached the thermo-electric element, 

 which by turning the rod could be caused to press against the 

 heated bar. Two thin mres, one of iron, the other of German 

 silver, formed the thermo-electric element, the wires being cut 

 across obliquely aud thus soldered together. The moveable rod 

 was graduated, and a fixed index was attached to the eud of the 

 cylinder ; it was thus in the experimenter's power to cause the 

 thermo-electric element to press against the bar at any required 

 distance from its heated end. The points of the bar thus ex- 

 amined were at a constant distance of two inches apart. By 

 means of a tube which communicated with the interior of the 

 glass cylinder, the latter could be connected with an air-pump 

 and exhausted. Duriug the experiments the entire cylinder 

 was immersed in a vessel of water, which was preserved at a con- 

 stant temperature of 12° C. 



The following tables contain the results of the experiments 

 with the substances therein mentioned. To facilitate compa- 

 rison, the temperature of the zero-point of each bar is reduced to 

 100 by multiplying each entire series of observations by a suit- 

 able factor. The figures under x represent the distances reckoned 

 from zero, and taken, as before remarked, from two to two 

 inches ; the adjacent figures represent the corresponding tempe- 

 ratures, and the last column in each case contains the mean of 

 experiments made with different bars. 



