Range of Meteorological and Magnetic Phenomena. 59 



mean for the ten first days of the year and that for the ten 

 last days, and to interpolate this difference over the days of 

 the year. The secular change has been taken into conside- 

 ration in no other way; so that the figures from which these 

 curves are drawn are simply the mean of the scale-readings 

 for a certain number of years. Except for my purpose, they 

 have therefore no value whatever. 



The scales on which these six curves are drawn are also 

 purely arbitrary. For H and Z even the scale-divisions 

 as published in the annals of the Koninhlijk Nederlandsch 

 meteorologisch Instituut at Utrecht have not been reduced to 

 any of the customary magnetic units. The factors by which 

 the original figures have been multiplied in order to form the 

 ordinates of these curves, ranging from 4 to J, were simply 

 chosen so that the consecutive maxima and minima of the 

 different curves should be, on an average, approximative^ of 

 the same importance. In other words, in the vertical dimen- 

 sions the curves have been compressed or expanded so as to 

 render the phenomenon I wish to show as prominent as 

 possible. 



And I think it will be admitted that it is prominent. Of a 

 few doubtful points I will speak later on. But if we except 

 these for the present, I think that it is perfectly apparent that, 

 however dissimilar the general behaviour of the individual 

 curves may be, every single maximum and every minimum in 

 one curve is, with an astonishing regularity, repeated in all 

 the others. Far from it that the curves should be parallel. 

 In many cases a maximum in one curve occurs earlier or later 

 than in most of the others. In still more cases, what is a 

 large bump or a deep valley in one curve may find its corre- 

 spondent in a hardly perceptible movement in another curve. 

 A few of these differences in behaviour shall be, later on, 

 explained away in a satisfactory manner. Others may be due 

 to insufficience, as yet, of the data at command. But even 

 if a certain — decidedly small — number of instances should 

 remain where one of the curves really has a secondary maximum 

 and accompanying minimum not shown by the others, why 

 should it not ? It may be a local anomaly; it may be that 

 indeed for, say, the barometer an agent enters also upon the 

 scene which has less grip, or none at all, upon the thermometer 

 or the magnetometers. There are certainly not many such 

 exceptions to the general phenomenon shown in the curves 

 under discussion. Indeed, they are so few that I incline to 

 think that we may predict even now that as soon as we shall 

 be possessed of a sufficient number of sufficiently good obser- 

 vations in every case it will indeed be found that to every 



