482 



marily the electron does not entirely return to the metal atom, but 

 that inversely a complete separation of the valency-electron from 

 the metal rest should be taken into account. It is this process that 

 we shall subject to a fuller examination in what follows, and which 

 we shall briefly denote by the expression of "primary reaction". ') 



Hence we come to realize the possibility of the conclusion that 

 the dissociation of the metal oxide does not take place by the 

 direct process of splitting up MO ^ M -1- 0, but that with increase 

 of energy first an activation sets in, manifesting itself as formation 

 of ions. From this activated condition the further course of the 

 reaction takes place. 



Thus we are led to place ourselves on the basis of these theories 

 of reaction-velocities which have appeared to possess remarkable 

 validity, at least in definite cases; particularly we come to the 

 "activation" basis given as of general validity first by Sv. Arrhenius') 

 and later among others by J. Perrin ') in his "Lnmière et Matière". 



d. lonisation Potential. 



It is known that the lonisation potential of the vapour is a decisive 



quantity for the detaching of an electron from a free metal atom. 



Where there is reason to assume that the primary reaction takes 



place in the gasphase*) we will first of all try, in connection with 



the view about the primary reaction given under c, in how far 



the lonisation potential can also govern the cyanizing-reaction. For 



V 

 this purpose we will calculate the values of the quantity — according 



to the table below, in which V represents the lonisation potential 



') A possible addition of the separated electron to the oxygen rest should be 

 considered as a second stage. From J. Franck's researches on the collisions of slow 

 electrons in electro-posilive and noble gases we know that negative electrons are 

 seized by oxygen, but are on the contrary under certain conditions relinquished 

 by the electro-positive atom. In the same way a partially bound electron, which 

 gets free through "critical" energy supply with small velocity [thus the lower limit 

 of energy supply required to bring about the primary reaction may be indicated 

 for brevity], may be bound to the oxygen atom. We leave this out of account in 

 the next close examination of the primary reaction. 



») Sv. Arrhenius. Zeitschr. f. Phys. Chem. 4, 226 (1889). 



5) J. Perrin. Ann. d. Phys. 11. 5 (1919). 



*) All the metal carbonates or oxides of the alkalies and earth-alkalies are 

 greatly or appreciably volatile at the indicated reaction temperatures. In the 

 dissociable carbonates the evaporation is promoted by the formation of oxydes in 

 molecular distribution (formation "in statu nascendi" and transportation by the 

 gas current). The parallelism between the cyanizing temperature and volatility of 

 the carbonates resp. oxydes is striking. 



I 

 1 



