i>r. Gregory's strictures ox don rodrigukz. 261 



notwithstanding, doubt the observations made with a stable 

 instrument by the English ? And let him not forget, that what- 

 ever error was thus occasioned in the distanoe between Bois- 

 commun and Chatillon, is more than doubledln all the remain- 

 ing triangles of the series, by reason of the bad shape of the 

 triangle, Chatillon, Boiscommun, Chateauneuf. 



If no error in the English observations can be fairly imputed N T o error ran 

 to the manner of fixing the zenith sector, neither can any be 2JJX^J ta 

 ascribed to the " construction" of the instrument itself. This the cmwtrttc- 

 was most positively declared by two very excellent judges, the t,on ^. the 

 late astronomer royal, and the Hon. Henry Cavendish, on their 

 close examination of the instrument. It will also be inferred, 

 without hesitation, by all competent judges, on reading the 

 description of it in the Phil. Trans, for 1603. To those who 

 have seen neither the instrument nor the description, it may 

 suffice, if I remark, that the equality of the divisions on the or to suUse- 

 arch, is evinced from this consideration, that on running the qnentderange- 

 micrometer screw from division to division, over the whole 

 arch, there was no where an indication of an error amounting 

 to half a second ; and that the instrument still continues free 

 from important " derangement," is tolerably well proved by 

 this, that the line of collimatton has been constant during all 

 the observations and all the journeyings of the sector, and 

 that it still continues the same. 



la the next place, it may be remarked, that no error in 

 observation can be imputed to a deviation from " vertical posi- or to a devia- 

 tion" in the sector. Important inaccuracy, in this respect, t !°" f rom . T er " 

 is precluded by the great length of the axis, by which the 

 instrument is rectified ; and by the ready and certain means 

 of placing the plumb-line directly over the illuminated dot 

 which marks the middle of the axis, or true centre of the 

 divided arch. For want of these admirable modes of correc- 

 tion, all previous instruments are necessarily imperfect. It 

 appears from Phil. Trans, for 1803, pp. 405, 406, t»at 

 when the instrument is adjusted in one position by means, of 

 the plumb-line and dot, it is turned to a position at right angles 

 to the former, and the adjustment confirmed ; and this being 

 the case in these two situations, the instrument must neces- 

 sarily be vertical in all others. 



Various reasons may be assigned to show that the sector could 



not, 



