624 PROCEEDINGS OF THE AMERICAN ACADEMY. 



be negative, as in iron, but tresfaihle compared with that in platinum. 

 Battelli,^'' stimulated by the fact that Haga had found a Thomson 

 effect in lead, undertook to measure this effect. He found it 2Msitive, 

 and gives data from which we get, — 



V at 53.0° C. = 4.37 X lO"'", 



V " 108.4° C. = 4.29 X " . 



Below 50° C. and above 110° C. we have, apparently, no direct 

 experimental evidence concerning the magnitude of the Thomson 

 effect in lead, and between these two temperatures we are, finding 

 Haga on one side and Battelli on the other, in doubt as to the sign 

 of this effect in the purest lead. It seems probable that the lead line 

 on the thermo-electric diagram should be very nearly horizontal below 

 50° and above 110° ; but it can hardly be said that we have a satis- 

 factory assurance in regard to this fundamental matter. Accordingly 

 it seems desirable to study very carefully through a rather wide range 

 of temperature the Thomson effect in some metal which is easily 

 obtained in a state of great purity. Lead is such a metal, but it is 

 mechanically difficult to work with, especially at moderately high 

 temperatures. Copper seems to be, on the whole, a better material for 

 the most careful direct examination, with a view to the establishment 

 of a reference line on the thermo-electric diagram. 



Appendix. 



The doubt concerning the amount and character of the lateral escape 

 of heat through the loose asbestos fibre packing was so great that two 

 of us^^ undertook in July, 1906, further experiments bearing on this 

 question. By means of a rough test with a small apparatus we satis- 

 fied ourselves that closely packed asbestos would not make a satisfactory 

 wrapping, as the flow of heat through it appeared to be even greater 

 than through the loosely packed asbestos which we had used. Ac- 

 cordingly we looked for other materials capable of bearing a tempera- 

 ture of 218° without injury, which might serve better. Magnesium 

 carbonate, which is used commercially as a bad conductor of heat, was 

 found to lose much weight, doubtless from the giving up of water of 

 composition, at 218° and to regain weight when exposed to the air 

 at ordinary temperature. This, therefore, was rejected. The oxide of 



" Atti della Reale Accadeniia dei Lincei, serie 4, t. 3, 1887, p. 61. Battelli 

 does not give tlie analysis of the lead, but states that its density at 0° C. was 

 11.344. 



18 Mr. Churcliill and myself. — E. H. 11. 



