228 Prof. B. Stewart and Mr. W. L. Carpenter. [Feb. 25, 



have used the words at first sight, because, apart altogether from the 

 comparatively small number of the Inequalities discussed, there is a 

 strictly terrestrial consideration which we must not lose sight of. 



It is well known to all magneiicians that we have not as yet arrived 

 at any wholly satisfactory method of separating between the disturbed 

 and the undisturbed magnetic observations, and the results now exhi- 

 bited have unquestionably been deduced from observations which 

 include a good many disturbances. Now under these circumstances 

 the effects of disturbances would only disappear from our results on 

 the hypothesis that such effects have no reference whatever to the 

 periodicities of which we have been treating — that they are, in fact, 

 non-periodic — so that they will become eliminated in a sufficiently 

 extensive series of observations. But we have much reason to sup- 

 pose that this is not the case, for the observations of Professor Loomis 

 and of Mr. John Allan Broun would seem to indicate that short- 

 period Inequalities of sun-spots occasiou terrestrial magnetic dis- 

 turbances, which follow closely on the celestial phenomena, so that a 

 maximum of sun-spots is quickly followed by a maximum of dis- 

 turbance. Now in the preceding tables we have discussed some of 

 the most prominent solar Inequalities in connexion with their mag- 

 netic effects, and doubtless the result we have obtained is a composite 

 one, its components being an Inequality of solar diurnal declination- 

 range (undisturbed), and an Inequality of disturbance declination- 

 range. We may add that Toronto is a station where the disturbance 

 is great, and also that the sun-spot Inequalities around 26 days are 

 greater than those around 24 days. 



Attempted Elimination of Disturbances. 



9. All these considerations point to the necessity of eliminating as 

 much as possible the effect of disturbances before we venture to 

 discuss our results. We have attempted to do this in the following 

 manner : — 



First of all, we would remind the reader that the Inequalities 

 around 24 and 26 days that we have been dealing with are most 

 probably not all the Inequalities around these periods, but only the 

 larger specimens of them. 



We remarked in our previous communication that observations 

 founded on sun-spots might present the same variety of period, when 

 treated as we have treated them, which they presented when 

 treated in another way by Carrington, who found that the spots in one 

 solar latitude had a different period of rotation from those in another. 

 If there be any truth in this remark, we might expect that the few 

 solar Inequalities which we have exhibited are only the most promi- 

 nent members of a comparatively large series, packed, it may be, so 



