Coherence and Recoherence. 169 
each other electrostatically, and also causing the resistance of 
the contact to fall. 
It is quite possible that there may be complications, 7. e. 
two or more of the above effects may occur at the same time; 
thus both adhesion and ionization may occur and conspire to 
produce coherence. 
One point must be borne in mind in these contact pheno- 
mena, viz., the part played by the films on the contact sur- 
faces. The contact between the two metal surfaces cannot 
pass until these films have been penetrated. The researches 
of Blondel*, Van Gulikt, Sundorpht, Guthe & Trow- 
bridge §, Shaw |, all show this point clearly. 
It is not easy to say, in a general case, how much of these 
effects is due to each of the following: (a) condensed air, 
(b) condensed water-vapour, (c) oxide on the metal. For 
this reason, an uncleaned surface often behaves in a most 
contradictory way. If the movable surface meets the other 
impulsively, the films are penetrated and a current passes ; 
whereas a gentle approach results in the films remaining 
intact and no current passes; and even if the surfaces be 
pushed together by reversing the current in the coil (fig. 2), 
it requires a great steady force to bring the metals into 
contact. In one case, using the balance apparatus, the force 
was measured, and found to be about 20 dynes. This is a 
very large force, since $ dyne is the force with which the 
surfaces are usually pushed together. It is hard to believe 
that any film other than a solid one could offer such resistance. 
Again, the voltmeter reading is not so directly related to the 
strength of coherence as would be expected if metal and no 
film existed at the contact. 
By scraping and drying the surfaces we can get rid of the 
films of condensed water-vapour and of oxide, and it is then 
found that, for several minutes at least, the effects are quite 
simple. 
In order to assist us in deciding between the rival theories, 
we will give two crucial experiments. 
(a) Suppose the two contact surfaces, in either of the above 
forms of apparatus, are brought together gently so that the 
films are not penetrated by an impulsive blew : no current 
will pass. When a spark is produced near them, coherence 
occurs: this shows that the surfaces are not more than about 
* Eelairage Flectrique, 1892. + Wied. Ann. 1898. 
1 Wied. Ann. 1899. § Phys. Rev. July 1900, p. 22. 
| Phil. Mag. March 1901. 
