PART I. ON MAGNETIC STORMS. CHAP. I. 



59 



Iii the table below, the scale-values appear as they were given us direct, as also the length of 

 one hour upon the original magnetograms. 



TABLE VIII. 



(') On Sept. soth the value was 4.38 and increased o.oi per diem up to Oct. sist, after which it was constant up to Nov. 25*. 



For convenience in the Plates, the sign is here fixed as follows: 



(,,, io d H u and e, are indicated by -f, when a deflection upwards answers respectively to increasing 

 II. I., increasing westerly declination, or increasing numerical value of V. I. 



The reduction of the magnetograms has been effected by a pantograph belonging to the Geogra- 

 phical Survey of Norway. The reduction to equal hour-length, and also the drawings, have been 

 executed by a very skillful cartographer, Mr. J. NATRUD of the Geographical Survey. 



As mentioned in my circular of June, 1903, it was my original intention to publish some of the 

 magnetic records by means of photographic reproduction. This mode of procedure, however, has 

 proved to be very unsuitable for the arrangement of curves from so large a number of observatories; 

 but I think that the method of reproduction chosen by us will be of equal value to science. 



