THE ELECTRON— COMPTON 211 



About that time Millilvan, who has always had a flair for picldng 

 strategically important subjects to which to devote his investigative 

 talents, undertook with his students a revaluation of the electronic 

 charge. Sources of error in the fog method were weU recognized: 

 Fog droplets were not all the same size, though measurements could 

 only be made on those smallest ones which fell most slowly; also 

 droplets did not remain of constant size, smaller ones tending to 

 evaporate and larger ones to grow; also there were unavoidable con- 

 vection currents in the air wliich modified the rate of fall of the fog; 

 and some droplets might contain more than one ion. 



Millikan cleverly avoided or minimized these difficulties by using 

 only a single droplet of some relatively nonvolatile liquid like oil or 

 mercury. By ionizing the surrounding air in an electric field he could 

 put various electric charges on the drop. Illuminating it by a powerful 

 light and viewing it like a star through a measuring telescope, he could 

 measure its rate of fall under gravity and its rate of rise when pulled 

 upward against gravity by an electric field, and keep repeating these 

 observations for hours. These measurements were so precise that, to 

 keep pace with them, he had to measure the viscosity of air with hither- 

 to unequalled accuracy. When all this was done, he had proved 

 conclusively that all electric charges are integral multiples of a funda- 

 mental unit charge, the electron, the value of which he set as 4.774 

 XIO"^" electrostatic units — about 40 percent larger than the earlier 

 estimates and believed by Millilvan to be correct within one part in a 

 thousand. 



Within the past half-dozen years, however, doubt has been thrown 

 on the estimated accuracy of this value from quite a different direction, 

 in work with X-rays. Originally, X-ray diffraction experiments in 

 crystals proved the geometric arrangement of atoms in the crystals, 

 but did not establish the scale of distances between atoms or the X-ray 

 wave length. These distances, once the arrangement of atoms was 

 known, were calculated from absolute values of the weights of the 

 atoms, which in turn were derived from electrochemical equivalents 

 and the value of the electronic charge. Thus X-ray wave lengths, 

 masses of atoms and distances between atoms in crystals all had values 

 dependent on knowledge of the charge of the electron. 



Recently, however, A. H. Compton, Bearden, and others have 

 succeeded in making measurements of X-ray wave lengths by diffract- 

 ing X-rays from a grating ruled with 15,000-30,000 parallel fine lines to 

 the inch, and operating near the angle of grazing incidence. These 

 measurements involve only knowledge of the number of lines per inch 

 on the grating, and the angles of incidence and diffraction of the X- 

 rays — both depending only on measurements of length and capable of 

 high precision. X-ray wave lengths thus measured were a little dif- 

 ferent from the earlier accepted values, and this cast doubt on the 



