200 Gen, Roy’s Account of a 
pote, or any other northern ftar, too little, and confequently,a 
latitude too great. Until triangles fhall have been extended 
beyond the Pyrenees, and the feCtor placed on the fouth fide of 
the range, the quantity of this attraction (by its double or 
counter-effect) cannot poffibly be afcertained. I will, however, 
only fuppofe it to have been io 7/ 8 7// to be deducted from the 
latitude of Perpignan, which will then become 42 0 4.1' 52' 
only three feconds lefs than that affgned to it in M. Cassini’s 
two books before mentioned. Thus the arc between Rodes 
and Perpignan will be i°39 7 21" 36 7// , and the total celeftial 
arc between Greenwich and Perpignan will be 8° 46' 48 /7 , as 
may be feen by attending to the four columns towards the left- 
hand of the annexed table of comparifon. 
With regard to the correfponding terrefirial arc, under 
which head are arranged the eleven columns towards the right- 
hand of the table, it is to be obferved, that various mealure- 
ments have at different times been made in different latitudes of 
the lengths of the degrees of the meridian, for the purpofe of 
obtaining, within certain limits at leaf:, the true figure and 
dimenfions of the earth. The mofi: effential operations of this 
fort, as having been executed with mofi: care, with the beft 
inftruments, and at the greateft diftances from each other, 
have all been done within thefelafr forty or fifty years; namely, 
in Peru under the equator, in middle latitudes in France and 
Italy, and in Lapland near the polar circle. The attraction of 
mountains, and unavoidable errors in the execution, will ever 
prevent juft conclufions from being drawn from the comparifon 
ot meaturements made too near each other. Thefe lafi: will 
always be found to differ more or lefs among themfelves. 
Sometimes even the refults may become abfurd or contradictory. 
In cafes of this fort, a mean of feveral fhould no doubt be 
5 taken 
