479 
Trigonometrical Survey. 
within a quarter of a minute of the truth; since the instrument 
was capable of being used with great accuracy, the arch hav- 
ing been divided by one of Mr. Ramsden's best workmen, and 
the value of one, or any number of revolutions of the micro- 
meter-screw, had been accurately obtained. If, therefore, any 
considerable errors have taken place in this part of the opera- 
tion, they must have arisen from the axis of the transit-instru- 
ment and the line on the cross board not being of the same 
height from the brass heads on which they were placed : but 
we think there is almost a certainty that this difference was 
confined, to such limits as will not introduce any errors of con- 
sequence ; for even supposing the register-heads were placed 
on the pickets so unskilfully that it became necessary to turn 
the screws on the levelling-board as much as they were ca- 
pable of, whilst the third remained unmoved, in order to adjust 
the transit, the error introduced on that account would be only 
half a minute, even though the hypotenuse should consist of 
but one chain, and be inclined to the horizon eight degrees. 
We therefore think ourselves justified in the opinion which we 
entertain of these angles being determined with sufficient ac- 
curacy ; since, if an error of one minute had taken place in the 
inclination of each hypotenuse, and those errors lay all one 
way, the length of the base, as hereafter given, would only be 
varied three inches by that circumstance. 
It may, perhaps, be imagined that some small errors have 
arisen from the handle of the chain not lying flat upon the 
brass heads when the new directions have been commenced. 
To obviate this, precautions were always taken to drive the 
pickets at the termination of the hypotenuses in such a man- 
ner, that the arrow on the handle could be made to coincide 
3 Q 2 
