264* Mr. Prideaux's Experimental Contributions 



below 200°, the latter much above that point : and in this 

 thermo-electric change of the same metal, at such manageable 

 temperatures, the secret seems to lie within narrow compass : 

 it promises to contain the key by which the whole field may 

 be laid open. Zinc bars were more troublesome, some being 

 always negative to others, whichever is heated. A long bar, 

 well cast, broken in the middle, and the contiguous ends used 

 for the points of contact, generally act consistently. But even 

 here the surfaces of contact must be filed, not hammered flat; 

 for a few blows of the hammer destroy the consistency of ac- 

 tion. The crystalline texture is pretty evidently concerned in 

 this inconstancy. 



Two slips of rolled zinc, cut side by side, from the same 

 sheet, and the heads doubled, or trebled, to retain the heat, 

 have always given me consistent results, so far as I have been 

 able to understand them. 



But it should be observed, that in this and other metals 

 (except lead) cast rods are more active than rolled or wire- 

 drawn pieces. 



23. Antimony is still more embarrassing than zinc. It can- 

 not be rolled or drawn, and is difficult to cast well. A tolera- 

 bly well cast bar, broken in the middle, and used as described 

 with zinc, was consistent as to direction of the current, but far 

 otherwise as to its force, when the pieces were consecutively 

 heated to the same degree, and applied one hot to the other 

 cold. When heated together, one half was always positive to 

 the other, as though the other had been of a less active metal. 

 The method which I found to answer best, was, to take as 

 many well cast bars as I could obtain, and classify them, by 

 heating them together in pairs ; putting the negative bars on 

 one side, and the positive on the other ; then, proceeding in 

 the same manner, to separate the negative bars from each 

 other ; and so on, until a pair was obtained which gave no de- 

 flection when heated together. Such bars gave consistent re- 

 sults when heated separately, the current being always thermo- 

 positive, in about the same degree as that of iron. 



It is singular that bismuth, a very crystallizable metal, has 

 never given me any embarrassment when acting in this way, 

 although sufficiently capricious in single masses (26.). 

 24. Thus the metals resolve themselves into two classes: 



1. Thermopositive *; those which become increasingly po- 

 sitive by heating. — Antimony, iron, zinc above 250°. 



2. Thermo-negative*; those which heat renders increas- 



* These terms are not unobjectionable ; but * idio-positive' and * idio- 

 negative' would seem to apply rather to the cold than the hot bar; and 

 * thermo-electro-positive' is too complex for an English word. 



