concerned in the Phenomena of ordinary Electricity, ^c. 315 



distinguished from instances of thermo-electricity by the circum- 

 stance of fusion and combination. 



There is an experiment of Schonbein which seems very difficult 

 to reconcile to the electrical hypothesis. He says, that if some 

 yellow solution of cobalt be poured into a glass tube bent into 

 the form of the letter U, and if a platinum mre, proceeding from 

 the galvanometer, be immersed in each limb of the tube, it will 

 be found that on heating the liquid in one of the limbs until it 

 becomes blue, a stream of electricity will move from the cold to 

 the heated column of Uquid, the strength of the current increa- 

 sing with the difference of temperature between the two limbs. In 

 one case Schonbein obtained a deflection of 40^. He obtained 

 similar results by heating certain acid solutions ; and he showed 

 that these are not cases of thermo-electricity, as might at first 

 view be supposed, but that the effects are attributable to the che- 

 mical change occasioned in one of the columns of liquid by the 

 action of heat*. Many experiments of this kind have since been 

 made by others. 



In the experiment with solution of cobalt, it is very hard to 

 conceive how a cui-rent of electricity could be generated in two 

 parts of one homogeneous, uninterrupted, and excellent liquid 

 conductor. The chemical action described by Schonbein is in- 

 ternal ; it is not exerted by the liquid on a second substance, but 

 on itself, within itself; the platina wii'es are mere conductors. 

 The conditions deemed necessary for electrical disturbance are 

 not present. The experiment, indeed, might be adduced as a 

 case wherein mere chemical action, without any electricity, pro- 

 duced deflection. But if electricity did act, it could scarcely be 

 of any other kind than that described in the beginning of this 

 essay, consisting of diff"erent constituents, the deflecting element 

 greatly predominating. 



In connexion ^^^th this experiment of Schonbein, it should be 

 recollected that the charge of CoUadon's Leyden battery of 4000 

 square inches, could only produce an average deflection of 20° or 

 30°, unless with an intense condensation of electricity, and then 

 he obtained but 40°, the same as Schonbein. Now, if Faraday's 

 law be applied to this case, the quantity of electricity in CoUa- 

 don's and Schonbein's experiments must have been the same. 

 Can imagination assist us in conceiving, or reason warrant us in 

 believing, that when a small tube containing a little solution of 

 cobalt is heated at one end, a quantity of electricity passes 

 through it equal to the most intense charge of a Leyden batteiy 

 of 4000 square inches, imperceptible to all else except the gal- 

 vanometer, and acting on that but feebly ? 



In my experiments, it is equally difficult to conceive that such 

 * Poggendorff's Annalen, 1838, vol. iii. p. 270. 



