306 Mr. E. H. Hall on the new Action of 



the current, and by E / the difference of potential of two points 



a centimetre apart in a direction at right angles to that of 



the current, while M has the same signification as before, 



we may write, as a very rough estimate for the case of iron, 



- E' 



= — =— = iQooooo a while for tin the value of this ratio may be 



as small as tgcTo Th^o oo* We may therefore conclude that the 

 ^equipotential lines in the case of static induction in glass, if 

 affected, at all by the magnet, are affected much less than the 

 equipotential lines in the case of a current in iron ; but Ave 

 cannot say that any such possible action in glass has been 

 shown to be smaller than the analogous action in the case of a 

 current in tin. 



I now go on with an account of further investigation of the 

 phenomenon actually discovered, and already in some measure 

 described in my previous article. When writing that article, 



E 



it seemed to me instructive to deduce the ratio, ^7, of the dif- 

 ference of potential per centim. on the longitudinal axis of the 

 gold-leaf strip to that per centim. on the transverse axis. 

 There were thus obtained, for the experiments made, values of 



E 



^7, ranging from 3000 to 6500, according to the strength of 



the magnetic field*. 



E xl 

 At that time I supposed that the ratio -^ — would prove 



to be a constant, not only for different strips of one metal, but 

 for all conductors. Subsequent experiments showed that this 

 was not the case ; and in this article the results obtained will 



. Mx V 



be expressed by the ratio — ^ — , where W has the same sig- 

 nification as before, while M now expresses the strength of the 

 magnetic field in cm.-grm. -sec. units, and Y= ~f the strength 



of the primary current divided by the area of section of the 

 conductor. This ratio does not prove to be the same constant 



* In obtaining this latter quantity, which was called M, a serious error 

 was made, and the value given was probably not much more than half 

 what it should have been. This fact was mentioned in a note when the 

 article in question was republished in Silliman's Journal for March 1880, 

 pp. 200 & 235. _ 



t This quantity V may be said to bear an intimate relation to the ab- 

 solute velocity of the electricity ; for if we were to take as the unit velocity 

 of electricity that of a unit current flowing through a conductor of unit 



C 



cross section, the velocity in any particular case would be a quantity -q- 



