Prof. B. Stewart and W. Dodgson. 



[Nov. 20. 



It will be noticed that the sums 96820 and 93739 are not far 

 different from those of the temperature inequalities given in Table III, 

 a result which would appear to indicate that for the purposes of this 

 inquiry it is better not to eliminate disturbances. 



16. In a paper puplished in the " Proceedings of the Royal Society " 

 (No. 192, 1879), by Balfour Stewart and Morisabro Hiraoka, certain 

 evidence is given that variations of magnetic declination, or what we 

 may call magnetic weather changes, occur later in point of time at 

 Tre van drum, in India, than at Kew, the difference being 9' 7 days. 



We are thus led to ask whether in our present investigation the cor- 

 responding phases may not be somewhat later at Prague than at Kew. 

 Taking Kew as our standard, and applying the test already fully 

 described in Article 6 of our present communication, we obtain the 

 following result : — 



Table XII. — Algebraic sum of Kew and Prague Magnetic Inequalities. 



Kew and Prague (Prague 1 division to right) .... =170005 



(together) =178078 



„ (Prague 1 division to left) =178501 



,, (Prague 2 divisions to left) .... =174147 



,, (Prague 3 divisions to left) .... =165505 



It would appear from this table that the phases of the various mag- 

 netic inequalities occur at Kew nearly one day before the advent of 

 the corresponding phases at Prague. 



In combining together the results of the two observatories, let us 

 therefore push the Prague inequalities one division to the left. 



17. When this is done, we obtain the following table: — 



Table XIII. — In which Kew and Prague are added together as now 

 described and the sums divided by 2. 



Division 



Magnitude of 



Division 



Magnitude of 



*om normal. 



Inequality. 



from normal. 



Inequality. 



-7-0 ... 



1123 



o-o 



. . 4866 



-6-5 ... 



3042 



+ 0-5 ... 



4618 



-6 -0 



3631 



+i-o ... 



. . . 4116 



-5-5 . 



. 3707 



+ 1'5 ... 



3528 



-5-0 ... 



2979 



+ 2-0 ... 



3102 



-4-5 , 



2111 



+ 2-5 .... 



■ 3501 



-4-0 



849 



+ 3-0 ... 



. . , 5459 



-3 -5 



3590 



+ 3-5 . , 



6229 



-3'0 ... 



. . 3304 



+ 4-0 ., 



, , , 4663 



-2-5 



3289 



+ 4-5 ... 



2651 



-2-0 



3017 



+ 5*0 ... 



1189 



-1-5 ... 



1690 



+ 5-5 . . 1 



2944 



-1-0 ... 



3168 



+ 6-0 ... 



. . . 2586 



— '9 ... 



. . . 4298 







