SUN AND THE EARTH'S MAGNETIC FIELD^ — FLEMING 185 



changes hour to hour from midnight to midnight, that is, from one 

 lovrer transit of the sun to the next, the data are studied for 24 

 hmar hours, that is, for the kmar day from one lower transit of the 

 moon to the next — an average of 2'3 hours 10 minutes for 24 lunar 

 hours. Wliile this variation is too small to have practical interest 

 in navigation, it is very important for the investigation of the high 

 layers of the atmosphere. The double wave of the lunar variation 

 indicates its tidal origin. It is likely that the lunar variations 

 originate in even higher layers of the atmosphere than the solar 

 variations, because they are so extrem|ely sensitive to changes in 

 magnetic activity. These geomagnetic tides provide a new approach 

 to the study of resonance phenomena in atmospheric oscillations. 



The minute annual variation is revealed when monthly values of 

 the magnetic elements are corrected for progressive secular change. 

 It should not be confused with the annual change which is the change 

 in one year due to secular variation. 



There are other magnetic time changes of quasi-regular and of 

 irregular type. The more important of these in relation to the sun are 

 short-period, long-period, and sudden-commencement disturbances 

 superimposed on the systematic variations and designated as magnetic 

 storms and perturbations. 



These are associated with what we call the earth's magnetic activity 

 or, let us say, its general state of magnetic rest or magnetic unrest, 

 and herein we find spectacular features emphasizing apparent rela- 

 tions to the observed solar phenomena. The magnetic activity at a 

 given station, during any interval, may be defined as a measure for 

 the frequency and intensity of marked irregular departures from the 

 normal diurnal variation in that interval. There are several ways 

 of applying such a definition. The simplest is that in which every 

 observatory, from inspection of its photographic records, assigns to 

 each 24 hours bet^veen successive Greenwich midnights a "character 

 figure," designated as 6', on a scale of 0-1-2. The character "0" 

 applies to quiet, "1" to moderately disturbed, and "2" to greatly dis- 

 turbed days. From these the International Association of Terrestrial 

 Magnetism and Electricity also determines the five least disturbed 

 and the five most disturbed days in eacli month and selects certain 

 days for rejDroduction in the annual publication of magnetic results 

 from each observatory. This measure, maintained by international 

 agreement since 1906, gives adopted average daily values for all 

 collaborating observatories — now numbering nearly 60. 



A more detailed measure is that depending upon the day-to-day 

 changes in the mean values of the horizontal component of the mag- 

 netic force since magnetic disturbances have a systematic effect on 

 that component. This effect is an initial rise, followed by a larger 

 decrease accompanied by more or less violent perturbations, and a 



