ircumstance not surprising with
regard to those regions at that time of year.

The eclipse of January 3, 1908, passed across the Pacific Ocean. Only
two small coral islands--Hull Island in the Phoenix Group, and Flint
Island about 400 miles north of Tahiti--lay in the track. Two
expeditions set out to observe it, _i.e._ a combined American party from
the Lick Observatory and the Smithsonian Institution of Washington, and
a private one from England under Mr. F.K. McClean. As Hull Island
afforded few facilities, both parties installed their instruments on
Flint Island, although it was very little better. The duration of the
total phase was fairly long--about four minutes, and the sun very
favourably placed, being nearly overhead. Heavy rain and clouds,
however, marred observation during the first minute of totality, but the
remaining three minutes were successfully utilised, good photographs of
the corona being obtained.

The next few years to come are unfortunately by no means favourable
from the point of view of the eclipse observer. An eclipse will take
place on June 17, 1909, the track stretching from Greenland across the
North Polar regions into Siberia. The geographical situation is,
however, a very awkward one, and totality will be extremely short--only
six seconds in Greenland and twenty-three seconds in Siberia.

The eclipse of May 9, 1910, will be visible in Tasmania. Totality will
last so long as four minutes, but the sun will be at the time much too
low in the sky for good observation.

The eclipse of the following year, April 28, 1911, will also be
confined, roughly speaking, to the same quarter of the earth, the track
passing across the old convict settlement of Norfolk Island, and then
out into the Pacific.

The eclipse of April 17, 1912, will stretch from Portugal, through
France and Belgium into North Germany. It will, however, be of
practically no service to astronomy. Totality, for instance, will last
for only three seconds in Portugal; and, though Paris lies in the
central track, the eclipse, which begins as barely total, will have
changed into an _annular_ one by the time it passes over that city.

The first really favourable eclipse in the near future will be that of
August 21, 1914. Its track will stretch from Greenland across Norway,
Sweden, and Russia. This eclipse is a return, after one saros, of the
eclipse of August 9, 1896.

The last solar eclipse which we will touch upon is that predi