324 EEPOET — 1887. 



method, and since I showed it to be impossible to derive exact results 

 from observations taken only hourly and not simultaneously, it being- 

 necessary to have as far as possible simultaneous observations by photo- 

 graphy, or, of course, by assiduous observations, as now was the case. 



As soon as the Pawlowsk observations were published I applied this 

 method to them, though the disturbances were only taken from the sup- 

 posed normals found by Professor Wild, a method which is also liable to 

 some arbitrariness. 



I arranged the disturbances thus found after their size, and found what 

 is exhibited in Table II. 



Dr. van der Stok, the Director of the Observatory of Batavia, who 

 was then at Utrecht, spoke with me about this method, and he extended 

 it somewhat more fully, submitting it also to the opinion of the members 

 of the Polar Committee. He did more, and calculated the whole of the 

 Batavia observations in this manner and sent me a proof-sheet of his 

 sixth volume, where, pp. 188-190, Tables XXXIII.-XXXV., are to be 

 found all disturbances distributed according to their size and sign for 

 each hour when they occurred. 



I have the honour to submit to you the result which I draw fx'om the 

 declination-disturbances alone, leaving it to hitn to attend to the ratio of 

 the easterly and westerly, which I saw to be nearly unity, and limiting 

 myself only to the distribution of the disturbances of each size over the 

 various houi's of the day. Table I. speaks for itself. When in this 

 manner the observations of a greater number of places will be discussed, 

 we shall be able to inquire in what manner the times of maximum and 

 minimum differ for various places. The first thing to be done appears 

 to agree concerning the method to be adopted — whether that of Chambers, 

 Whipple, or Wild, as mentioned in your report, or that of Dr. van der 

 Stok, combined with mine. 



I take this opportunity for calling your attention to the observations 

 of Utrecht (a place which is perhaps too near to Lisbon), but particularly- 

 to those of Batavia. If Utrecht be too near, I suppose, nevertheless, that 

 Mr. Schuster will admit that the situation of Batavia and the excellence 

 of the observations at that station require that it should be mentioned. 



Appendix II. Letter from Dr. Buys-Ballot to the Secretary. 



Utrecht : Api-il 23, 1887. 



Since you showed so much interest in determining the limits of the 

 Tariations of declination which separate those which occur more frequently 

 about noon or about midnight, I communicate to you the result of my 

 research after the variations at Jan Mayen Island. 



In the volume edited by Lieutenant Gratzl there occurs a table giving 

 the frequency of the deviations from 0-5, 5-10, &c., for the different 

 hours of the day. 



I asked him if he could not supply, from the original paper, the 

 frequency of the variations from 0-1, 1-2, &c., separately, and he has 

 kindly answered my request. 



The westerly and easterly deviations do not show a material difierence ; 

 only for the westerly deviations the limit seems to be between 5 and 6, 

 and for the easterly between 7 and 8. 



