ON COMPARING AND REDUCING MAGNETIC OBSERVATIONS. 327 



Perhaps they would still better agree if the deviations had been taken 

 from the true mean, and not from the, according to the method of Mr. 

 "Wild, somewhat arbitrary mean. It will be sufficient if I give here the 

 sum of the positive and negative deviations for every six Jiours which 

 show the greatest difference ; so I find : — 



1176 1375 1507 951 



a greater frequency from 5-16 local time, and for each of the other 

 differences without exception a greater frequency at the other hours. 

 Therefore I mention only the sum of all the deviations greater than 6 : — ■ 



23-4 5-10 11-16 17-22 



6+ 829 630 497 1050 



You will see that the range begins to be less pronounced for the 

 deviations greater than four minutes, the two species of disturbances 

 being already mixed. 



The International Polar Commission as a whole has declined to enter 

 into the proposal of Dr. Schmidt. I suppose, however, that every one of 

 its members, as much as I, is very desirous that regular magnetic obser- 

 vations might be furnished to Dr. Schmidt in order that he may continue 

 the work done by Professor Adams. — Yours faithfully, Buys-Ballot. 



Appendix III. Preliminary List of Magnetic Observatories. 

 By General Sir J. Henry Lefeoy and Mr. G. M. Whipple. 



The co-operative arrangements of 1839 were a great advance upon 

 any previous international arrangement for scientific purposes. They 

 wanted but little of completeness ; unfortunately that little was of essential 

 importance to the realisation of the object in view. There were no means 

 provided for promptly interchanging observations, and no uniformity was 

 established in the publication or in the scales adopted for measuring the 

 variations of the differeht elements in terrestrial magnetism. The result 

 has been an accumulation of volumes of observations, chiefly in quarto, 

 at widely distributed stations, scarcely any two of which can be directly 

 compared, the numerical values given requiring previous reduction to 

 common units. It is practically much the same as if meteorologists all 

 used arbitrary and different thermometers. But this chaotic condition 

 of the elements of our magnetical knowledge cannot be much longer 

 endured. The progress of the science is requiring more and more that 

 aU the material available for the elucidation of each class of phenomena — 

 those which depend on the earth's diurnal rotation directly, that is to say, 

 on the action of the sun, and those more remotely referable to it, or 

 perhaps resulting from causes independent of it — be brought together. 

 The following table has been compiled to facilitate such comparison by 

 showing what records exist, and where they are to be found. 



