1910.] TERRESTRIAL MAGNETIC DISTURBANCES. 139 



such that a complete circuit of the earth, // made, would require 

 from 3 to 4 minutes. In this connection it is worth while to record 

 that, judging from Dr. van Bemmelen's observations at Batavia, 

 Java, this interval of 3 to 4 minutes is also the average duration in 

 general of the " starting impulse " at any one station. 



After an interval of 3 or 4 minutes, the current either dies out 

 or gets so far away as not to produce any effect. But it may also 

 develop into a steady current continuing for an hour or more and 

 then break out anew into a second starting impulse and finally 

 develop into a complex current system with principal and secondary 

 current vortices, as shown by the more complicated cases of mag- 

 netic disturbances. 



Professor E. E. Barnard, a member of this society, has recently 

 published'' for the period 1902-09 a most valuable series of auroral 

 observations made at the Yerkes Observatory from which we quote 

 as follows : 



The streamers which spring from the arch as a base, and which always 

 have a decided lateral motion and last for a minute or so only, almost always 

 move to the west. On several occasions, however, I have seen them divide the 

 arch, with respect to their motion, so that the ones to the west moved west 

 and those to the east moved east. This is very rare. The motion is about 

 2° in one minute (and not 2 minutes to the degree as I stated in Astro- 

 physical Journal, vol. 16, 143). I have wished to determine this motion more 

 accurately, but we have had so few ray-producing auroras in late years, and 

 the rays are so transient, that I have not been able to do so. It would be 

 interesting to know if this motion is constant in a streamer and for all 

 streamers. 



The pulsating bright masses that usually appear in the northeast or 

 northwest, but which are sometimes seen under the pole, are among the most 

 interesting phenomena. They are sometimes present when there are no 

 other evidences of an aurora. 



If the motion of the streamers is at the rate of 2 degrees in one 

 minute, it would take five to ten minutes or more for the motion to 

 traverse a band from ten to twenty degrees wide. Now the possi- 

 bility of such a slow motion is contrary to what the cathode ray 

 theory with charged particles moving on the order of 60,000 miles 

 a second would premise, but it is in strict accordance with the 

 results announced regarding the non-simultaneity of magnetic dis- 

 turbances and their rate of progression over the earth. 



* Astrophys. Jour., Chicago, 111., Vol. 31, 1910, 208-233. 



