Atom and the Charge of Electricity carried by it, 525 



coefficients, so that the a's will not exceed a smal] number of 

 volts. To estimate the V's is more difficult, but we may 

 remark that when we produce a spark through a gas, in 

 which case there is strong evidence that we split up some of 

 the molecules into atoms, then, no matter how short the spark 

 | may be, or what may be the pressure of the gas, the potential 

 difference between the electrodes must exceed a certain value 

 which is very large compared with the potential differences 

 developed by the contact of heterogeneous substances, amount- 

 ing in the case of hydrogen to between 190 and 200 volts. 

 This minimum potential difference required to produce a 

 spark is so constant under very varying physical conditions, 

 such as pressure, spark-length, and so on, as to suggest 

 that it represents some property of the molecule ; and I am 

 ( inclined to think, and some experiments recently made at the 

 | Cavendish laboratory seem strongly to support the view, that 

 the potential difference between the atoms in a molecule 

 ] placed so as to be free from the action of other molecules is 

 j of the order of the minimum potential difference required to 

 produce a spark. In the few cases where a direct estimate 

 has been made of the work required to split up the molecule 

 into atoms, such as that made by E. Wiedemann of the work 

 required to decompose the hydrogen molecule, and that by 

 Boltzmann for the iodine molecule, the potential difference 

 indicated by these estimates far exceeds that produced by the 

 contact of heterogeneous substances. 



It would thus appear that in the case of gases where the 

 molecules are free, the condition 



2(*ci-*h) >Vi + V 2 



j is not fulfilled ; so that, on the electrical theory, chemical 

 j combination w T ould not proceed. To produce chemical com- 

 bination in such cases there must be some means of lowering 

 the potential difference between the atoms in the molecules. 

 Two methods by which this might take place at once sug- 

 gest themselves. The first of these is that the combination, 

 instead of taking place between a single pair of molecules, 

 really takes place between aggregates of the molecules, phy- 

 sical aggregation preceding chemical combination. 



Thus suppose that a number of molecules form themselves 

 into a chain, such as that represented in fig. 5 ; then, if we 

 consider a — H atom and the adjacent +C1 one, we see that 

 the disposition of the charges on the atoms of the other mole- 

 cules in the chain will diminish the work required to separate 

 the — charge on the H atom from the + charge on the 

 Phil. Mag. S. 5. Vol. 40. No. 247. Dec. 1895. 2 



