402 
ME. WAEEEN DE LA ELE ON THE 
carefully from them, the continuations of the images of these wires are depicted in posi- 
tions diametrically opposite. The wires are not central with the picture of the lunar 
disk, which was adjusted by means of the finder to be as much on the right of the plate 
as possible, in order that the details on the eastern limb might not be lost. It will be 
perceived that the light was inflected sufficiently during the taking of the picture to com- 
plete the image of the protuberance G, notwithstanding the intervening wire. 
About eighty seconds were required for covering and taking out photograph 25, placing 
plate No. 26 in the heliograph, drawing back the slide which covered it, allowing time for 
the vibrations imparted to the instrument to cease, and removing the temporary cover 
fi-om the telescope; so that the exposure of plate No. 26 commenced about two minutes 
and twenty seconds after the commencement of total obscuration, and continued until 
within a second of the reappearance of the sun, having been, like plate No. 25, exposed 
as nearly as possible one minute to the actinic influence of the prominences. The action 
of the prominences was, however, on account of the accidental disturbance of the heho- 
graph, not allowed to continue on one part of the plate during the whole time; and 
hence their impressions are not so strongly depicted as those of the prominences on pho- 
tograph No. 25, represented in Plate IX. fig. 2. 
On this account the appearance of photograph No. 26 v/as not easy at flrst to com- 
prehend ; and it gave me considerable difficulty for some time to make out uith precision 
the true nature of the result obtained. A gust of wind had arisen close upon the time 
this picture was being taken, and I was induced to imagine that the telescope had so 
been caused to vibrate ; but further examination of the picture showed that this could 
not have been the case, for three very distinct images were imprinted of the prominence 
E (the boomerang), proving that after each disturbance the instrument followed the 
sun’s movement correctly. This afforded a clue to what had actually occurred ; and I 
found on inquiry that two of my assistants had looked at the eclipse through the finder 
of the heliograph, and had, as it appeared, inadvertently disturbed it in right ascension, 
which the wear of the worm-wheel and tangent-screw permitted them to do. Fortunately 
a firm radius-bar, which I had had made previous to leaving England, held the telescope 
so firmly in declination that it could not be readily moved in that dhection ; if it had 
been so moved, the resulting picture might have defied interpretation, and have rendered 
the photograph useless for exact measurements. 
The clue once obtained, it was easy to make out three impressions of the wheatsheaf A 
(Plate XV.), three of the floating cloud D, three of each of the three points //, A", Ij!" of 
the fallen tree H, and less easily three'of the mountain-peak P because they overlapped. 
There were produced only two impressions of prominence Q on the western side, because 
at the epoch of the first action, when the point I was as yet partly visible, the moon still 
covered Q. The next impression was formed just when a part of Q had become wsible, 
and when also part of the spade L had been revealed. The prominence L appears to have 
been sufficiently brilliant to imprint itself continuously while its image was traversing 
the plate. In the last impression on the plate, the prominences and corona continued 
