475 
the parallax of a. Lyres. 
vations only + o'', 25. The observations of the same day 
give it -f- o",54 in favour of parallax. 
From June, 1813, to February, 1814, the telescope re- 
mained fixed, and six microscopes were used, so that the 
most uniform results might be expected. But we shall find 
the mean of 22 observations in June and July is half a second 
less than a mean of 17 in August. As six microscopes were 
used, the errors of reading must have been absolutely no- 
thing. The same may be nearly said of the bisections of the 
stars. The observations were made within half an hour of 
each other, and the arc intercepted, between y Draconis in 
the zenith, and a Lyras, was less than 13 0 . All these cir- 
cumstances would have led us to expect, provided there were 
no parallax, an agreement to less than T *- of a second. 
This induced me to make further examination of the ob- 
servations of this year, and I found by 61 days, from June to 
December inclusive, in which both stars were observed, and 
for which the reductions are given by Mr. Pond, 
the constant of parallax = -f- o",42. 
The circumstance I have now to mention is remarkable 
Mr. Pond considers the interval between the beginning of 
July and the 14th of November, as sufficient for, and favour- 
able to this enquiry : I therefore omitted the last 5 of the 
preceding 61 days of observation, and then found the 
constant of parallax = -f- o",89, really differing little 
from my parallax. 
I next reduced the only 5 double observations in January 
and February, 1814, and added them to the former 61 re- 
duced by Mr. Pond, and now found 
the constant of parallax = -f- o",i8 ; 
