204 ANNUAL KEPORT SMITHSONIAN INSTITUTION, 19 3 5 



whereas these rays uniformly illuminate the region to the right. If 

 no rays, other than those of the curvatures represented in the 

 diagram, were present in the primary radiation, the observer would 

 find a sharp cut-off in the intensity at 6. By measuring this cut-off 

 angle, knowing the strength and extent of the field, the sign of 

 charge and the rigidity can be determined. If rays are bent in the 

 opposite direction, as indicated by the dotted orbit, the region of 

 low intensity lies on the opposite side of the vertical. Thus, both 

 the rigidity and the sign of charge are determined by the angle 6. 



In case the radiation were not of a single rigidity, but had some 

 kind of a distribution over all values, the sharp cut-off would be 

 replaced by a gradually changing intensity. The difference between 

 intensities at two angles would be due to rays whose rigidities lie 

 within the range between the two cut-off values. If positives and 

 negatives were both present, the intensity difference at two angles 

 would be equal to the excess in the number of charged rays of one 

 sign over the other. 



Thus the angular measurements determine the distribution with 

 respect to rigidity of the excess of one sign of charged ray over the 

 other. To resolve the distribution of each sign separately angular 

 intensity measurements must be combined with results obtained by 

 varying the position along the meridian line O" . In the case of 

 a single value of the rigidity, the intensity from any direction such 

 as B' would remain uniform, as the observer moves along the merid- 

 ian, until the position 0' is reached where the intensity from this 

 direction would fall suddenly to zero. In the case of a distribu- 

 tion this sudden drop would also be replaced by a gradually chang- 

 ing intensity and, in this case, the difference between intensities in 

 any two positions is due to rays of both signs of charge, in propor- 

 tion to their numbers, in the ranges of rigidity determined by the 

 cut-off angle Q' at the two positions. Thus, this type of measure- 

 ment determines the distribution with respect to rigidity of the sum 

 of the positive and negative rays together. Combining this result 

 with the distribution of the excess of positives over negatives, we 

 can discover the distribution of each component separately. 



A change of intensity with position along the meridian can also 

 be recorded by an apparatus which measures intensities from all 

 directions, for example the electroscope, but the analysis of the dis- 

 tribution from this type of measurement is not as straightforward 

 for the reason that the gradual change of intensity due to the chang- 

 ing angle of cut-off cannot be distinguished from the changing in- 

 tensity which takes place at each angle due to the distribution 

 of rigidities. 



