150 Mr. Wollaston's Description 
judge, seemed always to deviate equally, perplexed me much ; 
and destroyed all confidence. The collimation in altitude 
(whose error when constant, is unimportant) appeared va- 
riable ; and seemed to give uncertain conclusions. Whether 
that took its rise from the object-glass, or the wires ; from the 
hanging of the plumb-line, or the microscopes ; was doubtful. 
All these things it took up much time to investigate. But I 
think I now may say, these difficulties are all surmounted. 
The error in the Y s, it has been said already, is cured. The 
object-glass I suspected might have some little shake, from its 
being attached to a false tube on the outside, and therefore 
liable to be touched, instead of being within the tube of the 
telescope itself. This was made more certainly steady. The 
wires I was sure did not move. Neither did the microscopes, 
after I had set them perfectly at ease ; for I found that in the 
first placing, I had a little cramped one of them. The dots 
came then to be examined, by which the plumb-line is ad- 
justed ; or rather, I should say, by which the circle is brought 
into a position for adjusting the microscopes. Here a small 
error was discovered, ft has been mentioned already, that 
there are four pair of dots for this purpose. Though laid off, 
I am satisfied, at first with the greatest care, and strictly true; 
the opening or enlarging of them afterwards, to make them 
just visible on each side of the plumb-line, had occasioned 
some very small differences in them, in respect of their ad- 
joining divisions or dots on the limb. The adjusting there- 
fore to a different pair of these dots, which I had done, would 
necessarily occasion a difference in the collimation. This be- 
ing avoided, by using always the same ; and other causes of 
error being removed ; the collimation for altitude seems now 
