AURORA AND MAGNETIC DISTURBANCES 
OF AUGUST 27, 1916: - 
Wy TEE, miference to the aurora reported by Mr. 
: W. FE. Denning at. Bristol on August 27, be- 
tween 2 and 4 a.m. G.M.T. (Nature, August 31, 
_ p- 551), the Director of the Meteorological Office notes 
a report received from Mr. J. Ernest Grubb, observer 
at Seskin, near Carrick-on-Suir, Co. Waterford, Ire- 
land, that aurora was visible there on the night of 
August 26 between 10.5 and 10.40 p.m. G.M.T. The 
greatest display noted by Mr. Grubb occurred about 
10.25 p.m., when streamers from N.W. to N.N.E. 
stretched to within 20° or 30° of the zenith. The 
light was sufficiently brilliant to illuminate the in- 
terior of a room facing W.N.W. 
At Eskdalemuir, Dumfriesshire, in spite of a 
cloudy sky, auroral glow in the N.W. was reported 
by the observer at 9 p.m. on August 26 and 1 a.m. on 
August 27. The magnets at Eskdalemuir and Rich- 
mond (Kew Observatory) were considerably dis- 
turbed, especially at the former station. The full 
amplitude of the disturbance there in the north and 
vertical components cannot be assigned, owing to the 
limit of registration being passed, but the range in 
each element considerably exceeded 400 y (1 y=1x 10-* 
C.G.S.), and in the west component it was fully 
gooy. At Kew the ranges of the horizontal force 
and vertical force were approximately 250 y and 200 y 
respectively. Declination at Kew had a range of 
27', the most rapid movements occurring early on 
August 27; the extreme easterly position was reached 
at about 2.5 a.m., and the extreme westerly at about 
2.35 a.m. 
A notable feature, especially at Eskdalemuir, was 
the “‘sudden commencement,” introducing the storm 
at about 7.45 p.m. on August 26. Its oscillatory 
character was particularly well shown in the north 
component, a very rapid fall of 11 y preceding a rapid 
rise of 1007. After this commencement the horizon- 
tal component at Eskdalemuir and Richmond (Kew 
Observatory) ‘remained above its normal value until 
about 10 p.m., when it fell below normal and remained 
so, while oscillating considerably, during the rest of 
the disturbance. The depression in the horizontal 
component at Kew Observatory at 6 a.m. on August 
27, when the storm was nearly over, exceeded 100 y. 
An interesting feature in the vertical force curves at 
Eskdalemuir towards the end of the storm after 6 a.m. 
on August 27 is.a series of oscillations of short period, 
averaging about 4:6 minutes, which recall a similar 
phenomenon observed there in the storm of Novem- 
ber 5-6, 1915. 
Tue display of Aurora Borealis on August 26-27 was 
observed by Dr. John Satterly in Canada, Writing 
from Jackson’s Point, Lake Simcoe, Ontario, Dr. 
Satterly says that on August 26 the whole northern 
sky from horizon to zenith was illuminated for several 
hours. On the horizon there was a strong yellowish 
glow with streamers radiating upwards. Arcs of light 
encircled the zenith, and flickering bands and patches 
of colour were seen in middle altitudes. The smallest 
newspaper headlines could be read at tr p.m. On 
August 28 the northern lights were feeble, but at 
10.30 p-m. (Eastern time) an immense riband of light, 
practically a complete semicircle, spread across the 
sky. It extended from the east and rose a few degrees 
south of Jupiter, threaded Pegasus diagonally, cut 
Cygnus, passed through Lyra to the north of Vega, 
- and dipped down through Hercules to the west. Stars 
in their apparent rotation passed across it, so that the 
NATURE 
band was fixed relatively to the earth. The arc inter- | 
NO. 2447, VOL. 98] 
57 
sected the Milky Way at about 60° or 70°, very nearly 
at the zenith; it was much brighter and narrower and 
more definite than the Milky Way. No portion of the 
are appeared in the northern sky. At 11.15 p.m. the 
western half faded away gradually, and at 11.30 p.m. 
the eastern half vanished. 
LHE SAKURA-JIMA ERUPTION OF 
JANUARY, 1914. 
ROF,. OMORI has recently published a second 
valuable memoir on the eruption of Sakura-jima, 
which occurred on January 12, 1914.' The volcano 
lies in Kagoshima Bay, in South Japan, a few miles 
to the east of the city of Kagoshima. Until the last 
eruption Sakura-jima was an island. It is now con- 
nected by a lava-stream with the east side of the bay- 
The part of the bay lying to the north of Sakura-jima 
ranges in depth from 70 to 107 fathoms, and is 
apparently of the same origin as the deep lakes which 
are found behind the sea-coast volcanoes of Usu-san 
and Tarumai-san. 
Displacements of the Ground.—Soon after the erup- 
tion it was noticed that the sea-level had undergone 
an elevation relatively to the adjoining coast. At high 
tides the low districts at the ‘south end of 
Kagoshima were flooded. Along the north-west 
and north coasts of Kagoshima Bay the rise 
of the water was still greater, embankments 
and stone walls being damaged and _ extensive 
rice-fields devastated. The amount of the sea-level 
elevation was nearly a metre at Kagoshima and two: 
metres or more in Sakura-jima. The apparent sea- 
level elevation was greatest at the end of rg14, after 
which it began to decrease. 
Prof. Omori attributes this change to the depression 
of the ground in the neighbourhood of Sakura-jima in 
consequence of the great eruption. This depression 
was revealed with greater precision by a renewal of 
the Military Survey levelling of the district and of 
the levelling along the railway lines near Kagoshima 
Bay. Prof. Omori has represented the results on a 
map on which are drawn the curves of 50, 100, 300, 
and 500 mm. depression. The curves of 300 and 
500 mm. run close to the coast of the northern portion 
of Kagoshima Bay. The axes of these curves, which 
are directed north and west respectively, intersect in 
a point lying off the north coast of Sakura-jima, and 
Prof. Omori regards this point as indicating the centre 
of the area of greatest depression of the ground. The 
total depression-volume within the 1too-mm. curve ‘Is 
1:35 c. km. The aggregate volume of lava flows and 
pumice and ash ejection during the recent explosion 
is about 2:2 c. km. 
It is important to notice that the point of maximum 
depression, which probably coincides with the prin- 
cipal centre of the lava reservoir, lies, not under 
Sakura-jima, but in the region between it and the 
active volcano of Kirishima. 
The triangulation surveys of 1898 and 1914 also 
reveal considerable displacements, both horizontal and 
vertical, in Sakura-jima, while the coast of the island 
is everywhere depressed. Three points in the interior 
have been raised 0-14 and about 9-7 and 10-4 metres. 
The horizontal displacements in the north-west of ‘fhe 
island vary from 2-04 to 3-62 metres towards the south, 
and in the west and north by amounts from 
1-08 to 4-52 metres towards the north and north-east. 
The north and south portions of Sakura-jima have 
thus been displaced outwards in contrary directions. 
1 “The Sakura-jima Eruptions and Earthquakes." II. Bull. Imp 
Earthq. Inv. Com., vol. viii., 1916, pp. 35-179. The first memoir was noticed 
in Nature, vol. xciv., p. 289; see also vol, xcii., pp. 716-17. 
