IO 
NATURE 
[NovEMBER 5, 1903 
commenced, but, thanks to the magnificent work of 
Respighi, Tacchini, Ricco, and Mascari, we have prac- | 
tically a continuous record of them 
time. 
The question then arises, are these prominences in 
any way related to the occurrence of magnetic storms ? 
Before answering this, a few preliminary remarks 
may be made. 
In the first place the number of spots on the sun is 
nearly always insignificant compared to the number of 
prominences. Prominences are, therefore, of greater 
relative importance than spots. 
While sunspots are restricted to practically a narrow 
zone (+5° to +35°) on each side of the solar equator, 
prominences can and do occur all over the sun’s disc. 
Again, the general trend of the spot circulation is from 
the higher to the lower latitude, while in the case of 
prominence the reverse happens. In some years we 
have a great number of prominences near the solar 
equator, while in other years they are observed also in 
great numbers near the solar poles. A glance at some 
curves recently published in this Journal (vol. Ixviii. 
P. 257, July 16) will shown not only the general drift 
up to the present 
18600 18700 (8800 18900 
to 
CROSSINGS * 
WIDENED 3° 
LINES. © 
©. 
+00 
PROMINENCES. 
60°- 90° 100 
N. & 6. LAT 
(Tacchini) 
PROMINENCES. 
60°- 7o° Loo 
N.LAT 
(TACCHINL) 
DAYS re 
er 
MAGNETIC 
OISTURBANCESS 
“GREAT” 
(eccis) 
i8eo00 ‘8700 IsBO00 
Fic. 1.—Comparison showing days of “great” 
and crossings of widened lines. 
epochs of sun-spot maxima and minima respectively.) 
(890-0 
of prominence activity, but the epochs when the 
prominences were present near the polar regions of 
the sun. The years in which they attain these high 
latitudes are not numerous; they are (first observations 
made in 1870) 1870, 1871, 1881, 1882, 1892, 1893, and 
1894. Mr. William Ellis, who has made a special 
study of magnetic disturbances, has tabulated the num- 
ber of days of ‘‘ great’? disturbances, that is, those re- 
corded at Greenwich above a certain standard. These 
are utilised to form the following table :— 
Number of days 
Number of ae rH Average 
; a Yeats" pevictlisturbances PEF Yeat 
1870-1872 3 16 Ae : 
1873-1880 8 2 a 
1881-1882 3 10 g.38 
1883-1891 9 I oll 
1892-1894 3 7 233 
1895-1900 6 2 0°33 
Another and perhaps more 
the coincidence of the epochs 
of magnetic disturbances and polar prominences is 
illustrated in the accompanying figure (see NATURE, 
February 19, vol. Ixvii.). 
The continuous and broken vertical lines indicate the 
epochs of maxima and minima sunspots, showing that 
the former tend to occur later than the peaks of the 
magnetic curves. In this diagram Respighi’s obsery- 
NO. 1775, VOL. 69] 
striking way of showing 
of the occurrence of days 
magnetic disturbance, polar prominences 
(The continuous and broken vertical lines indicate the 
ations of polar prominences made during the years 
1870 and 1871 are not included, but their mean latitudes 
for each hemisphere during these years were +70°, 
It will thus be seen from the above that the occur- 
rence of polar prominences is closely associated, at any 
rate in time, with great magnetic storms, and, there- 
fore, with aurora, which nearly always accompany 
them. Further, prominences fulfil the three conditions 
mentioned previously in this article, for they can occur 
when there are spots and also when there are none. 
One argument used against the prominence theory — 
is that the polar prominences are “ quiet’’ prominences 
and therefore are not likely or are possibly not capable 
of producing such large terrestrial effects. The critic 
may, however, have. forgotten to consider the 
possible and natural conclusion that the appearance 
of prominences in high latitudes may at least be simply 
an indication of greater action occurring nearer the 
equatorial regions with consequent greater extension 
of the disturbed region towards the polar zones. 
That the polar regions of the sun are sometimes 
greatly disturbed is again emphasised by the presence 
of enormous streamers that are seen in those regions 
during some total solar eclipses. Fur- 
ther, these polar streamers are observed 
only at those times when the prom- 
inences approach high latitudes. Here 
again we have good cause to doubt the 
inability of these polar prominences as 
disturbing agents. Even if the prom- 
inences be not conceded to be the initial 
cause of magnetic storms, their gradual 
changes of position towards the solar 
poles may afford a valuable means of 
forecasting the epochs of magnetic dis- 
turbances. 
From the facts before us let us con- 
sider the question of forecasting the 
years in which magnetic storms should 
occur. If the reader will glance at the 
figure accompanying this article and 
continue the curves on the assumption 
that the last sunspot minimum occurred 
19000 
‘9000 
in Igo1-5 and the next maximum in 
1905, he will most probably make 
a maximum fall between these two 
dates, but somewhat nearer the latter; in fact the 
maximum would have been placed in the middle of the 
year 1903. It will be noticed, however, that at the 
sunspot maximum of 1870 the disturbance curve reaches 
a maximum a year after 1870. A recent investigation 
has indicated that all sunspot cycles are not alile in 
| intensity, and that the cycle commencing in Ig0I may 
probably correspond to that which commenced in 1867. 
If, therefore, the coming sunspot maximum. should 
attain the same dimensions as that reached in 1870, it 
seems quite possible that the magnetic disturbance 
curve for the present cycle should correspond to that 
portion commencing about 1867. If this be so, then not 
only should polar prominences be recorded from the 
years 1903 to 1906 or 1907, but during these years 
““ great ’? magnetic disturbance will be liable to occur. 
As shown in the previous table, no less than 16 of 
Ellis’s “ great”? magnetic storms occurred between 
1870 and 1872; also two occurred in 1869 and one in 
1868, so that if we consider the present year to 
correspond to 1868 there is much in store for us. It 
may be mentioned also that since the years 1899 and 
1900 the prominences have exhibited the tendency to 
attain high latitudes, so that there seems every reason 
to suppose that magnetic storms and aurorz may be 
experienced during the course of the next three or four 
years, after which there will be a cessation for about 
ten or eleven years. Witiiam J. S. Lockyer. 
