TOTAL SOLAR ECLIPSE OF JULY 18, 1860. 381 
counting whole seconds by two steps, when using a chronometer which beats half- 
seconds. 
Column 3 contains the computed errors of the chronometer for the several epochs, and 
Column 4 the epochs of the photographs, corrected by deducting the numbers in 
column 3 from those of column 2. 
The time-intervals in columns 5 and 6, the nature of which is sufficiently explained 
by their superscriptions, will be found useful in checking the numbers given in this and 
the subsequent Tables, they having been employed in the calculations. 
Columns 7 and 8 give the results of the measurements of the sun’s radius by the two 
methods already explained. On examining these numbers certain discrepancies will be 
apparent ; and they may, on the whole perhaps, appear at first sight not so accordant 
as might have been expected. The greatest difference in column 7 is between No. 29, in 
which the sun’s radius measured 1909-5 thousandths of an inch, and No. 16, in which 
it measured 1900-5 thousandths — the difference being y^^^ths of an inch, or about 
4" -5. Some of this difference is due to errors in centering ; for on taking the means 
of columns 7 and 8 for the photographs Nos. 8 and 16, the difference of radius was 
reduced to y^-oths, =3"-5. Measures of the same photograph may vary in difficult 
cases, on account of the irregularity or faintness of the sun’s border, from yx^th to 
xo^ o o ths of an inch, but in most cases they are in complete agreement. There is, how- 
ever, a real difference of photographic diameter in different pictures ; for in disturbed 
states of the atmosphere the sun’s diameter is enlarged by irradiation ; and, on the other 
hand, when once the instantaneous apparatus has been so adjusted as to produce the 
best effect, any great diminution in the intensity of the light renders the picture more 
feeble, and the periphery of the sun consequently less distinctly defined ; and it is just 
barely possible that the fainter portions of the limb are not depicted at all, whence 
would arise a diminution in the size of the picture. 
The mean of the measurements given in column 7 is 1904-17, and of those in 
column 8, 1905-65, — the difference not being greater when converted into arc than 0"-7, 
while the mean of both sets of numbers is 1904-91. By assuming the radius of the 
sun to be 15' 44''-8, as calculated by Mr. Hind from Leveeeiee’s Tables, the value of 
yoVoth of an inch of my scale becomes 0"-4960, the logarithm of which is 9-6954654. 
This number has been employed in the reduction of the several measurements to their 
equivalents in arc. 
Column 9 contains measurements of the moon, which are very accordant. The 
original negative of the second totality-picture presented greater difficulty in centering 
than the first totality-picture, in consequence of the triplication of the images of the 
protuberances ; a positive albumen copy of it on glass was more easily centered. The 
greatest discordance in the measures is 4-5 thousandths, which are equivalent to 2"-2. 
The mean of all the measures gives 2002-25 thousandths =16' 33"‘l as the radius of 
the moon, which agrees almost exactly with the number of Mr. Caeeington, 16' 33", 
and that of Mr. Faeley, 16' 32"-9. 
MDCCCLXII. 3 F 
