230 Dr. J. Monckman. Occluded Gases and [May 31, 



number of experiments with various bodies, among which is gas- 

 carbon. The hot junction was 100° 0., at which temperature the 

 deflection produced by a couple (carbon and copper) was negative, 

 the same as copper-platinum, but a little larger. He does not appear 

 to have worked at higher temperatures (' Annales de Chimie,' vol. 8, 

 1866, p. 415). 



Knott and MacGregor also worked with gas-carbon, and in 1879 

 published a paper in the ' Transactions of the Royal Society of 

 Edinburgh,' vol. 28, in which a line for carbon is given. The 

 material was in the form of a cylinder 15 cm. long, 1*5 cm. thick. A 

 strong heated wrought-iron tube, 4 inches long, 2 inches diameter, 

 and 1-inch bore, closed at one end, was suspended over the junction 

 and allowed to cool gradually. 



From 230° downwards the line is parallel to that of platinum. 

 Above 230° it appears somewhat uncertain ; they speak of it thus : — 

 " For a small range of temperature (to 230° C.) it is possible to 

 express the deflection in terms of the first and second powers of the 

 temperature, the following formula holding good : 8 = —8*29 -f 

 0-604 1 + 0-000385^; above 230° C. it does not, perhaps because of 

 chemical changes, produced by heat. Carbon appears to be an excep- 

 tion to the general law." " The above formula and the graphic 

 treatment enable us at a higher temperature to determine its posi- 

 tion " (see fig. 8). The position and slope of the lines are opposite to 

 those now used. 



Such a result did not appear to agree with the experiments 

 already described, and as I had found gas-carbon a very unsuitable 

 body for use where two pieces were required having anything like the 

 same thermo-electric power, it appeared probable that good results 

 might be got with the other rods ; and as carbon and platinum form 

 for 230° parallel lines I decided to use a couple consisting of these 

 two bodies. 



Nine series of observations were taken, using three different 

 methods, of which it will be sufficient to describe the last. 



Near one end of a carbon rod a hole, about 5 mm. in diameter, was 

 drilled, and into this the end of a platinum wire was inserted and 

 fixed by being wedged with a piece of rod carbon. The whole was 

 thoroughly covered with Indian ink, which, when dry, was again 



Fig. 10. 



