SOLAR ECLIPSE OF AUGUST 29, 1886. 
3.57 
photograplis of the solar lines, which had to be replaced by a scale divided in tenths 
of millimetres. 
The early morning of the 27th was cloudy and showery, and all practice impossible 
at the eclipse hour, but the Sun was observed later on in the day, and the position ot 
the various points of the limb accurately determined for all circumstances that might 
arise in the use of direct or inverted images, of solar prisms, or of projections. 
On August the 28th the sky was quite clear an hour before sunrise, and, observing 
the Sun at the time when first contact was to take place on the morrow, I found the 
definition very good, but the low altitude made the limb somewhat unsteady. I was 
then using the power of 110 on the Jones 4-inch equatorial. Changing my instru¬ 
ment, I then took a number of readings of the solar lines, using the first and second 
orders of the spectrum with the powers of 6^ and 4 respectively. As I again found 
the second order with power 4 to be much more distinct than the other combination, 
I resolved to adhere to my intention of adopting the second order for any observa¬ 
tions on the morrow. 
At 2 A.M. on the 20th, the morning of the eclipse, not a cloud was to be seen, and 
at 4 A.M. the stars were still shining brightly in every dii’ection, although a slight 
breeze had sprung up from the South. The wind then shifted gradually towards the 
East, and at 5 A.M. clouds were fast beginning to appear. Soon it became but too 
evident that rain was falling at no great distance to the North-East of our station, and 
heavy clouds began to show themselves in the direction of the lising Sun. External 
contact took' place in the midst of rain, and the first glimpse of the Moon was 
obtained through the Jones equatorial when one-third of the Sun’s surface was 
already obscured. The clouds then cleared off rapidly, and we could safely uncover 
our larger instruments. The sky, however, remained only fairly good until the near 
approach of totality. This was particularly unfortunate, as the interval between first 
contact and totaliiw would have been most valuable for testing the more delicate 
adjustments of our instruments, and for preliminary observations. 
As soon as the Sun’s image could be seen upon the cap of the slit. Lieutenant 
Helby placed the centre of the radial slit on the middle of the outer arc of the solar 
crescent, and kept it there as steadily as possible ; but no bright lines came into view 
between W.L. 5600 and h until one minute before totality, when the first line seen 
was 1474 K., which stood out very brightly, and then followed almost immediately a 
number of bright lines close by h, on the less refrangible side. I estimated their 
number at about fifteen. They seemed to be of different lengths, but I did not see 
them long enough to judge of their relative intensities. The height to which 
1474 K. extended from the photosphere might be about 8' of arc. The exact position 
of this line and the general position of the group were fixed by the lines of the solar 
spectrum, which had been under my eye for some time previous. I never moved the 
grating, or the viewing telescope, during tlie observations, so that everything was in 
excellent adjustment the whole time, and the field of view was well known. The 
