VOL. XLI.J PHILOSOPHICAL TRANSACTIONS. 417 



observations are greater than the correction itself. Mr. Celsius remarks fur- 

 ther, if the difference of latitude between Dunkirk and Paris be supposed to be 

 2° 12' 12V'. which is a mean between 4 others he mentions, the length of a 

 degree will amount to but 56,395 toises. And if the observations at Malvoi- 

 sine and Amiens, be counted according to Mr. Bradley's theory, for the inter- 

 val of a month between the observations, the length of a degree will come out 

 to be 56,926 toises : which is 135 toises less than the length of a degree, found 

 by measuring the whole length of France ; and 134 less than that of Mr. Picard, 

 so highly approved of by Mr. Cassini, as confirming his own. 



Mr. Celsius having finished his remarks on the observations made in the 

 north part of France, extending from Paris to Dunkirk, proceeds to examine 

 those taken in the south, from Paris to Collioure, near the borders of Spain, 

 and the Pyrenean mountains. By the former, a mean degree was found to 

 consist of 56,960 toises, by the latter 57,097; and consequently the earth is 

 an oblong spheroid. 



Mr. Celsius, in examining these observations, which were taken under the 

 conduct and direction of the late M. Cassini, in 17OO, first considers the struc- 

 ture and goodness of the instruments used ; then the accuracy of the astronomical 

 observations for finding the difference of latitude ; and, in the last place, the 

 trigonometrical operations for determining the distances of places ; especially 

 the two extremes under the same meridian. 



The principal instrument M. Cassini carried with him, was, a limb of 12 

 degrees, whose radius was indeed 10 feet, but divided only into degrees and 

 minutes ; the other parts were added to it at Perpignan. Here Mr. Celsius 

 observes, that the finding the true centre of this limb was, and still is, a very 

 difficult and troublesome problem to a good artist ; that no mention is made, 

 whether the position or place of this centre, and the divisions of the limb, were 

 ever examined at Paris or Collioure, though the carriage of the instrument 

 through so long and rough a way, could not but make some alteration in the 

 place of the centre. 



It is true, the zenith distance of Capella, taken by it at Paris, was confirmed 

 to be right by another instrument ; but it cannot be concluded that the zenith 

 distance of the same star, taken at Collioure by this instrument, and not con- 

 firmed there by another instrument, must be true also. For the point of di- 

 vision, answering to this distance in the limb, was not examined ; and a centre 

 wrong placed may by accident give the true zenith distance, viz. when the 

 true and erroneous centre happen to lie in the same perpendicular to the 

 horizon. 



The exceptions taken to the astronomical observations, for finding the dif- 



VOL. VIII. 3 H 



