the effects of parallax, &c. of certain fixed stars. 333 
the first and second magnitudes. The stars 7 and /3 Aquilae 
appear to have a parallax as great, or greater, than « Aquilse. 
The results of these observations are given in Table 3. 
It is to be remarked, that these results cannot be con- 
sidered nearly so exact as those of Table 1, because the 
observations are not nearly so numerous, and because the 
coefficients of p are in general much smaller. This latter 
circumstance could not be avoided in some of them, on ac- 
count of their being too faint to observe in strong day 
light. For some of these stars also, the number of ob- 
servations is so few, that a continuance of them may alter 
considerably the results ; but with respect to others, this is 
not the case. 
The value of p has not been deduced for « Aquarii, because 
of the smallness of its coefficients ; but as this star shows a 
much less discordance than the others, it would afford, as 
well as y Draconis, an argument favorable for the explana- 
tion by parallax, were not its zenith distance so great, that 
some uncertainty with respect to refraction may take place. 
The results, contained in the three first tables, have been 
deduced from so many observations, that it is impossible that 
the principal conclusions, although relative to such minute 
quantities, can be materially affected by the variable errors 
of observation. If error exist, it must be from some cause 
not to be controlled by mere observations. Two causes 
suggested themselves, which seemed to require particular 
consideration. 
1. The instrument being in such different states as to 
temperature in summer and winter, may, by changing its 
