380 
ME. WAEEEN DE LA EEE ON THE 
Although the differences of declination and right ascension of the cusps from the 
centres were taken out in every case, they will not be recorded in this report, because 
the results have been worked out by other measurements. I cite the above merely to 
show the applicability of the instrument to the measuring of two such coordinates as 
declination and right ascension. 
Measurements of the moon’s radius could, I found, be obtained with great accuracy 
by centering all the photographs which had been taken so near totality as to give a large 
proportion of the lunar periphery. Experience in centering for the sun had pre\iously 
proved that the operation could be performed with ease, even in those cases where a 
large portion of the sun’s disk was shut off by the moon. The totality-negatives were 
also measured, and 1 had two positive copies on albuminized glass made by superposing 
the negatives ; and these also were measured. In the case of the totality-pictm-es, 
besides the radius, the whole diameter was measured, by causing the slide A to carry 
the picture from one side to the other of the centre of the reading-microscope. These 
measurements are given in Table I. column 9. 
The readings, it will be observed, are in inches and decimals of an inch, the values of 
which in arc are based entirely on an assumed diameter for the sun, in consequence of 
no steps having been taken in Spain to obtain data for an absolute scale. This was not 
attempted, because at that time it was not contemplated that such an extensive use of 
the photographs would afterwards be made. It might be practicable to ascertain the 
value in arc of an inch on the screen of the heliograph ; but this would have no retro- 
spective application to the pictures obtained in Spain, because it would not be possible 
to place the object-glass, the secondary lens, and the screen in relative positions abso- 
lutely the- same *. By assuming for the semidiameter of the sun its tabular value, the 
arbitrary measurements were translated into measures of arc ; but ic will be seen that 
any error in the tabular semidiameter of the sun necessarily affects all the numbers based 
upon it. 
Table I. 
In Table I. are given the following quantities: — Column 1 contains the progressive 
numbers of the photographs. Where a break occurs in the consecutive order, it arises 
from the circumstance that the omitted plate was spoiled after the picture had been 
taken, or that the photograph could not be taken, for the reasons already given. 
Column 2 contains the times noted as the epochs of the photograph, to the nearest 
half-second. In observing the time, the half-second beats of the chronometer vrere 
counted, from I to 120, a method which, I think, is less likely to lead to error than the 
* Since this paper was handed in, the Kew heliograph has been removed to Crairford, and attempts have 
been made to obtain an absolute value in arc for the scale of the micrometer b_y procuring pictures of objects 
situated at the distance of about a mile. Hitherto these attempts have been attended with only partial 
success, on account of the want of definition of the resulting pictures, consequent on the feebleness of the 
light and the movements of the intervening atmosphere. I have it in contemplation to continue the expe- 
riments and to erect a scale of equal parts about 80 feet long, at a known distance, with the view of ascer- 
taining the radial distortion of the image, and the value of each increment from centre to cii’cumference. 
