of Electrification by Evaporation. 405 



Hence, even if the deflections cl of the tables be really due to 

 evaporation, and if evaporation be the principal source of the 

 atmospheric electricity, the quantity of water which would be 

 required to produce a single flash of lightning is very much 

 greater than the actual quantity ever found in a thunder- 

 cloud, while the thunder-cloud usually gives, not one. but 

 many flashes. 



It may be objected to this reasoning that a thunder-cloud 

 in some way collects the electricity obtained by the evapora- 

 tion of a much larger quantity of water than that contained 

 in itself. On this account it is of interest to consider the 

 total evaporation on the globe. If 100 centim. be taken as 

 the average annual rainfall for the whole surface of the earth, 

 and 400 centim. be taken as the value of A, the total annual 

 evaporation on the earth is sufficient to produce only about 

 14,000,000 flashes of lightning from clouds 9 square kilo- 

 metres in extent. Hence, if one were able to observe every 

 flash of lightning which occurs within 18 kilometres of himself, 

 the average number of lightning-flashes seen by such obser- 

 vers at different points of the earth's surface in one year would 

 be only twenty-eight. Calculation from more exact data would 

 give a less number. This calculation assumes that all the elec- 

 tricity is discharged through lightning- flashes, whereas a very 

 large portion of the atmospheric electricity is discharged in 

 other ways — e. g. in the aurora borealis and other atmospheric 

 phenomena which are referred to electricity, and in quiet dis- 

 charges which take place without any noticeable accompanying 

 phenomena. 



Evaporation is then, at most, a very insignificant source of 

 the atmospheric electricity. But, further, the following facts 

 are to be observed : — 



1. The deflections obtained in the experiments were alwavs 

 very small. (In the original readings it was easy to make an 

 error of 0*1 of a division : and though the numbers in the tables 

 are the means of several observations, they are still affected by 

 this error of reading.) 



2. The sign of the apparent charge is not always the same 

 for the same liquid. This is particularly noticeable in the im- 

 portant case of water. 



3. The deflections were very much diminished when care 

 was taken to eliminate very small sources of electricity. Still 

 the deflections obtained with the evaporating-dish dry, are 

 usually much larger than the difference between these and the 

 deflections obtained when a liquid was evaporating. 



Evidently, then, most of the electrification was clue to other 

 causes than evaporation; and the experiments do not certainly 



