712 Dr. S. J. Allen on a Null Instrument 



Puttin 



we have 



and there is instability if\k\ < tv and «<a 4 /x\/ (tt) — & 2 » 



The range o£ q is thus very small compared with that in 

 the leading case, and the influence within it comparatively 

 feeble. 



When fi^q is in the neighbourhood of 3 the range of 

 instability and rate of change of amplitude are of order a 6 ; 

 and so on, the order of smallness increasing with the 

 integral value to which fij/q approximates. 



August, 1907. 



LXXI. A Null Instrument for Measuring Ionization. By 

 S. J. Allen, Ph.D., Instructor in Physics in the University 

 of Cincinnati *. 



THE present paper is the outcome of an attempt by the 

 author to devise and construct an apparatus for the 

 measurement of ionization, particularly that caused by 

 radioactive substances, in which the sensitive quadrant 

 electrometer would be used as an indicating, and not as a 

 measuring instrument. The method used is quite simple, 

 and has proved thus far, in a set of preliminary experiments, 

 to be both convenient and accurate. An account of it may 

 therefore be of interest to those engaged in radioactive 

 research. 



The quadrant electrometer, until the advent of radio- 

 activity, was considered as an extremely difficult and incon- 

 venient instrument to use with any degree of accuracy, but 

 since the development of the Dolezalek type results may 

 now be obtained with it which are on the whole fairly 

 consistent. Its sensitiveness and quickness of reading make 

 it very convenient for use in radioactive work, especially in 

 those experiments where the activity of the substance is 

 rapidly changing. Nevertheless, even with the best instru- 

 ments, there is at times a tendency to sudden variations 

 generally traceable to static induction effects. In many 

 experiments it is extremely difficult to avoid these by 



* Communicated by the Author. 



