272 DE. CHARLES CHREE: SOME PHENOMENA OF 



includes H, V, and I, and so, when taken in conjunction with (C) and (C'), affords a 

 complete treatment of disturbance so far as it goes. 



The disturbed days dealt with in (C) see (C') IA numbered 209. 

 In selecting them regard was paid exclusively to the greater or less irregularity of 

 the movement, and not to its mere amplitude. In dealing with magnetically quiet 

 days at Kew, the practice has been to smooth the curves with a pencil so as to 

 replace small oscillations or irregularities by curves of continuous curvature following 

 the general trend of the undisturbed curve. This procedure was also followed in 

 dealing with ordinary day curves. The 209 highly disturbed days were really those 

 on which the declination curves were in part, or in whole, so irregular that it did not 

 seem possible to smooth them satisfactorily, and which had thus to be omitted from 

 the first essay at dealing with the regular diurnal inequality. There was no intention 

 originally of utilising them in any way for the deduction of a diurnal inequality. 



As \v;is explained in (C), the curves for different years were under consideration at 

 different times, so that no very uniform standard of disturbance probably existed. 

 A further important consideration is that as declination was the only element under 

 review in (C), declination curves only were consulted when the day was classified as 

 disturbed. 



During a magnetic disturbance one element is often much more disturbed than 

 another, and the times at which the disturbance appears greatest in the different 

 elements are usually different. Of the three elements D, H, and V, the last is 

 unquestionably that which usually is least -disturbed, while H is most. Thus, of the 

 209 days classed in (C) as disturbed, there are a number which would probably have 

 been treated as ordinary if V had been the element under review, and there are even 

 fair number whose H curves might probably have been smoothed. On the other 

 hand, there are a considerable number of days not included in the 209 which would 

 certainly have been classed as disturbed if H had been the element originally dealt 

 with. I decided, however, that arbitrary as the choice of the 209 days unquestionably 

 was, from the point of view of the H and V curves, it was best to adhere to it so as 

 to have all the disturbance results relating to a common set of days. 



On several days the Kew trace was defective or beyond the limits of registration 

 once or oftener during the 24 hours. Thanks to the kindness of Mr. E. KITTO, 

 Superintendent of the Falmouth Observatory, it was possible to fill up nearly all 

 these blanks fairly satisfactorily by reference to the Falmouth curves. On a few 

 occasions the limits of registration at Kew and Falmouth were exceeded at one 

 common hour during the 24. For these a reading was accepted answering to the 

 edge of the photographic sheet. The error so introduced is believed to have been so 

 small as to be practically without effect on the inequalities presently considered. 

 On one occasion, however, August 20, 1894, the limits of registration in the H sheets 

 at Kew and Falmouth were exceeded at several common hours, and it appeared best 

 to omit the day when dealing with H. On the few days of incomplete trace the 



8 



