124 DISSERTATION SECOND. [,ar T h. 



ally to its advancement. Nothing, indeed, is so hostile to 

 the interests of truth, as facts inaccurately observed 5 of 

 which we have a remarkable example in the measurement 

 of an arch of the meridian across France, from .Amiens to 

 Perpignan, though so large as to comprehend about seven 

 degrees, and though executed by Cassini, one of the first 

 astronomers in Europe. According to that measurement, 

 the degrees seemed to diminish on going from south to 

 north, each being less by about an 800th part than that 

 which immediately preceded it toward the south. From 

 this result, which is entirely erroneous, the conclusion first 

 deduced, was correct, the error in the reasoning, by a very 

 singular coincidence, having corrected the error in the 

 data from which it was deduced. Fontenelle argued that, 

 as the degrees diminished in length on going toward the 

 poles, the meridian must be less than the circumference 

 of the equator, and the earth, of course, swelled out in the 

 plane of that circle, agreeably to the facts that had been 

 observed concerning the retardation of the pendulum when 

 carried to the south. This, however, was the direct con- 

 trary of the conclusion which ought to have been drawn, 

 as was soon perceived by Cassini and by Fontenelle him- 

 self. The degrees growing less as they approached the pole, 

 was an indication of the curvature growing greater, or of 

 the longer axis of the meridian being the line that passed 

 through the poles, and that coincided with the axis of the 

 earth. The figure of the earth must, therefore, be that 

 of an oblong spheroid, or one formed by the revolution 

 of an ellipsis about its longer axis. This conclusion seem- 

 ed to be strengthened by the prolongation of the meridian 

 from Amiens northward to Dunkirk in 17*3, as the same 

 diminution was observed ; the medium length of the de- 

 gree between Paris and Dunkirk being 56970 toises, no 

 less than 137 less than the mean of the degrees toward 



