390 
MR. H. H. TURNER ON THE 
and the other lines (excepting A and i) are seen by Lockyer in spots. In the 
Egyptian eclipse— 
g and i .... were seen 7 minutes before totality ; 
k and 3 „ 
and all the other lines ,, „ 2 ,, ,, ,, 
In making my own observations, the slit of the spectroscope was placed radial to 
the limb at the cusp some minutes before totality. The point of the cusp was brought 
on to the slit, and the motion of the Sun allowed to carry it gradually away. The 
Solar spectrum, as seen in the spectroscope, thus became narrower as the point of the 
cusp approached, and finally disappeared. It was hoped to catch the bright lines just 
at this point, the dark lines of the vanishing Solar spectrum serving as a set of fiducial 
mai’ks for the identification of the bright lines. But there was also inserted in the 
focus of the eyepiece a photograph of this region of the spectrum taken the day before. 
Some twenty minutes before totality it was suspected that, in the increasing dark¬ 
ness, the lines of the photograph would not be well seen, and the places of the lines 
were therefore carefully marked by scratching the film with a sharp penknife ; the 
scratches were seen to accord well with the proper lines in the Solar spectrum. But 
in the actual observations it was not found necessary to refer to the photograph, and 
the identifications mentioned below were referred to the vanishing spectrum. For 
the period of the observations the cusp does not travel very rapidly along the limb,* 
so that the slit was nearly radial until a few seconds before totality. I now subjoin 
tlie observations, as recorded at the time by an assistant to my dictation. 
Time by chronometer. 
li. m. s. 
G 55 30 
G 5G 30 
7 0 20 
7 7 45 
7 8 55 
7 10 10 
7 1] 30 
7 12 0 
7 15 50 
7 20 50 
No bright line visible at cusp. 
Cloudy ; no bright line visible up to the time clouds 
appeared. 
Clear again. 
F line appeared. 
g appeared : very short. 
Cdoudy for ten seconds. 
g and i. 
Immediately after many lines appeared. 
Totality. 
End of totality ; cloudy. 
Only F, g, and i visible at times. 
* The rate of motion is very nearly inversely proportional to the duration of totality. Eor a three- 
minute eclipse the angular distances of the cusp from the point of disapjiearance are respectively 75°, 60°, 
51°, 33|°, at 44 mins., 90 secs., 54 secs., 18 secs, before totality. 
