472 
MESSES. EOSCOE AND THOEPE ON THE MEASITEEMENT OF THE 
accumulating. During the night of the 21st much rain fell ; but shortly after sunrise 
on the 22nd the clouds in great part disappeared, every trace of haze was dissipated, and 
the sky was of the purest blue : it will be seen that the measurements of photochemical 
action made up to the time of first contact are almost absolutely coincident with the 
mean numbers derived from the observations of the three preceding days. As the eclipse 
progressed, and the temperature of the air fell, clouds were again formed, and from l b 40 m 
up to the time of totality it was impossible to make any observations, as the sun was 
never unclouded for more than a few seconds at a time ; indeed just before totality a 
slight shower of rain fell. As the illuminated portion of the solar disk gradually increased 
after totality, the clouds rapidly disappeared, the estimated amount falling from 9 (over- 
cast =10) to 3 in about fifteen minutes. The observations were then regularly continued 
to within a few minutes of last contact. 
Although the disk and by far the greater portion of the heavens were completely obscured 
by clouds during the period of totality, rendering any determination of the photochemical 
action perfectly valueless in view of our special object, it was yet thought worth while 
to attempt to estimate the chemical intensity of the feebly diffused light at this time. 
That it has a certain degree of actinism is of course evident from the fact that photo- 
graphs can be taken during totality. Immediately, therefore, after the supposed com- 
mencement of totality the slit was opened, and the sensitive paper exposed for ninety-five 
seconds ; as the end of totality was very plainly indicated about two or three seconds 
after the slit had been again closed, it was clear that the paper had been exposed only 
during the period of totality. Not the slightest action, however, could be detected on 
the paper ; on comparing them in the dark room it appeared considerably lighter than 
the extreme light end of the graduated strip. The calibration of the fixed strip employed 
in our measurements was not carried out to the extreme limits of the tinted portion, but 
on the supposition that it decreased uniformly in shade from end to end ; or, in other 
words, assuming that the curve representing the decrement in intensity of tint maintained 
its symmetrical character throughout, which is a perfectly legitimate assumption to 
make when we bear in mind the manner in which these tinted strips are made, it follows 
that the amount of chemical action during the totality could not exceed 0003 of the 
unit which we adopt, and in all probability it is much less. 
The results of the day’s observations are contained in Table IV. 
