Theory of Nobili's Coloured Rhigs. 13 



thinks, inversely proportional to the Jirsl power of the radius, 

 but what is partly evident, even from simple inspection, to its 

 cube. 



It is not easy to comprehend how M. Becquerel's observa- 

 tions correspond with considerable accuracy to the first power. 

 It is however possible that a kind of polarization produced by 

 the electromotive reactionof the products of decomposition may 

 have been concerned in it. This cannot be supposed to be 

 owing to the oxygen, because, even where the plate is difficultly 

 oxidizafele, it is applied to the higher oxidation of the oxide 

 of lead ; but the unequal film of the peroxide of lead itself 

 may be taken into account, by which perhaps the nearer points 

 of the plate become negative to the more remote parts, which 

 may have produced an increased thickness of the film upon the 

 latter in comparison to the former. And in fact, e. g. copper, 

 which becomes oxidized when exposed to the influence of a 

 powerful current, became negative to that which formed the 

 positive electrode when a weaker current was used. 



M. Becquerel also calculated the form of the curve which 

 limits the film of the product of decomposition in that case 

 in which, instead of the mere point, a shorter or longer piece 

 AE of the wire formed the positive electrode. According to 

 him, this form would be given by the definite integral 



Adz 



in which m and % retain their former signification ; z represents 

 the perpendicular variable distance of the point P of the wire 

 from the plate, the action of which upon the point C should 

 be exactly observed, and n the distance EB (see the figure). 



But this formula is not really based upon true notions, be- 

 cause, e. g. dz, in the manner in which it enters into it, cannot 

 well represent anything more than a constant section perpen- 

 dicular to the direction of the current, whilst, when measured 

 at z, it forms a variable angle with this direction, which only 

 becomes a right angle when 2 = 0, and whilst partial currents 

 flow in a four-sided pyramid from the surface of the base to 

 the apex, yet every instant pass over a different section. Lastly, 

 that in this case in determining the partial path of the current, 

 the object has again been attempted in a perfectly arbitrary 

 manner, and with the omission of the most essential conditions, 

 after the above remarks, does not require further illustration. 



Dr. E. du Bois-Reymond. 



