1910] on Magnetic Storms. 773 



fixed mirror, and a curved line, the distance of any point of which 

 from the straight hne represents the corresponding angle between the 

 two mirrors. The light from the fixed mirror is interrupted for a 

 few minutes every hour or second hour, and the gaps thus produced in 

 the base line serve to mark the time. 



When a magnet, instead of having its motion limited like the 

 compass needle to the horizontal plane, is so supported at its centre as 

 to be perfectly free to turn, it dips, making with the horizon an angle 

 which varies from 0° in the magnetic equator to 90" at the magnetic 

 poles. In London at present the dip is about 67". The dipping 

 needle must coincide in direction with the resultant magnetic force 

 on either pole. This force has thus in general both a horizontal 

 and a vertical component, and the intensity of each is continually 

 varying. At a complete magnetic observatory there are three 

 magnetographs in constant action, recording respectively declination, 

 horizontal force, and vertical force. In the Kew pattern instrument 

 each magnetograph has a separate drum and a separate sheet of paper, 

 but the three drums are driven by a single clock. To economise 

 paper, two days' traces are usually taken on the same sheet, the 

 position of the light being shifted at the beginning of the second day 

 to avoid superposition of the traces. At Kew, the hour of changing 

 papers is shortly after 10 a.m. [Slide.] 



In some foreign types of magnetograph, e.g. the Eschenhagen, 

 which was used in the National Antarctic Expedition of 1001-4, 

 the three elements are recorded on one sheet of paper, but only one 

 day's record is taken on each sheet. [vSlide.] 



§ 8. In my subsequent remarks I am obliged to employ a term 

 which has unfortunately more than one meaning. It will be simplest 

 to explain these meanings l)y reference to the daily variations of 

 temperature which are familiar to everyone. Suppose that in the 

 present month of March we record the temperature at every hour 

 and take a mean value for each hour of the 21 from all days of the 

 mouth. We shall then find a regular progressive rise from a 

 minimum, probably at G a.m., to a maximum probably at 3 p.m., 

 and then a gradual fall to the minimum. The difference between 

 this maximum and minimum is known as the range of the regular 

 diurnal inequality for the month. On individual days, however, the 

 hours at wliich the highest and lowest temperatures occur will vary, 

 and if we take the mean of the dift'erences between the highest and 

 lowest temperatures of each individual day, irrespective of the hours 

 at w^hich they occur, we get a totally distinct range which I shall 

 call the mean absolute range. 



It is obvious on reflection that the absolute range cannot ])e less 

 and must usually be considerably greater than the range of the 

 regular diurnal inequality. At Kew, for instance, taking the four 

 years 1892-5, the mean absolute range of temperature for the year 

 as a whole was 13° * 9 F., while the range of the regular diurnal 



