348 TRANSACTIONS OF THE CANADIAN INSTITUTE. [VoL. VI. 
On the 2nd of September there certainly is a slight depression, the 
spot then being on the Eastern limb. On the 9th and roth the magnetic 
storm was very severe, and these were the days of the spots’ centrality. 
It passed off the disc on the 15th without any co-incident depression. 
On the 22nd it was central on the other side of the sun, and still no 
depression is to be seen. On the 28th it again appeared, by rotation, 
and a slight depression can just be noticed. It was once more central 
on our side—of course much reduced in size—on October 5th, but 
neither at that time, nor until its disappearance after the third rotation 
on November 8th, was there any magnetic effect traceable to it. This 
spot seems to have recurred a little in advance of the rotation period, 
which is about 27% days. It should, however, be observed that on 
August 17th, before it broke out, there was a noticeably sharp depres- 
sion, preceded by a slighter but still decided one on July 20th. Also 
that a dip took place on October 8th—these dates corresponding with 
the period of solar rotation. Does it not then begin to appear that the 
cause of the magnetic disturbance is not the spot, but some violent 
change, below the photosphere and in the body of the sun, which gives 
rise to the spot also? For this change, whatever may be its nature, 
affects the magnets here in advance of the appearance of a spot, as well 
as after it has vanished. 
This can be seen in dozens of cases, on even a casual examination of 
the magnetic curves, but one other instance only will be quoted here—a 
spot being taken for illustration which appeared nearer to the time 
of sun-spot maximum. The diagram shows the variations. of the 
JANUARY, 1894. FEBRUARY 
123456789 1011 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 3t 12345 
1st Solar 
rotation 
hea A 
I 
FEBRUARY MARCH bar 
6789 10 11 12 13 14 15 16 17 18,19, 20721 22 23 24 25 26127 28 1 [2 3 4'5.6 7 8§ 9 10 11 12 13 14f 
2nd Solar 
rotation 
-> 
a 
