240 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 1910. 



varying between the limits 1 and 150, and which were upon as diverse 

 substances as oil, mercury, and glycerine and found in every case the 

 original charge on the drop an exact multiple of the smallest charge 

 which we found that the drop caught from the air. The total number 

 of changes which Ave have observed would be between 1,000 and 2,000, 

 and in not one single instance has there been any change which did 

 not represent the advent upon the drop of one definite invariable 

 quantity of electricity, or a very small multiple of that quantity. 

 These observations are the justification for assertions 1 and 2 of the 

 introduction. 



For the sake of exhibiting in another way the multiple relation- 

 ship shown by the charges on a given drop the data of Table I have 

 been rearranged in the form shown in Table II. 



Tablk ii. 



No more exact or more consistent multiple relationship is found 

 in the data which the chemists have amassed on combining powers, 

 and upon which the atomic theory of matter rests, than is found in 

 Tables I to XIII. 



DIRECT OBSERVATION OF THE ENERGY OF AGITATION OF A MOLECULE. 



Before discussing assertion 4 it is desirable to direct attention to 

 three additional conclusions which can be drawn from Table I : 



1. Since the time of the drop in the field varied in these observa- 

 tions from 380 to 6.7 seconds, it will be seen that the resultant moving 

 force acting upon the drop was varied in the ratio 1 to 55, without 

 bringing to light the slightest indication of a dependence of e x upon 

 the velocity. Independently of theory, therefore, we can assert that 

 the velocity of this drop was strictly proportional to the moving 

 force. The certainty with which this conclusion can be drawn may 

 be seen from a consideration of the following numerical data. Al- 

 though we had upon our drop all possible multiples of the unit 

 4.917X10" 10 between 4 and 17, save only 15, there "is not a single 

 value of e x given in the table which differs by as much as 0.5 per cent 

 from the final mean e x . It is true that the observational error in a 



