■ ON THE MERIDIAN. " ... 



degrees ufed with the F7'^;2i:/i' degree, gives "that ratio as i: 1,0034295^, 

 Th^fecond mean of the fame three degrees ufed with the English de- 

 gree gives itas i: 1.0031913 ; and the third mean of the fame three 

 degrees ufed with the Swedish degree gives it as -i: 1.00324170,, 

 and the mean of there three, means gi^-es the.ratio of the- polar to the 

 equatorial diameter as 1: 1.0032895, or the compreffion at the poles 

 ■rsL. or '^i^jT very nearly: and this ratio has been finally adopted for 

 computing the general fcale of degrees both of iktitude'' and longitude, 

 and alfo of the degrees perpendicular to thejineridian^. from the equa* 

 tor to the pole.:. 



It is^ v/cU fcnov/n tO' mathematicians, that if a meridian of the earth 

 be an ellipfe, whatever may be the compreffion at the poles, the increr- 

 Tnents to- the firft degree of an aro on that meridian to. make it equal 

 to any other degree north from it, will alivays be as the increment to 

 the fquare of the latitude of that diftant d-gxee,. above the fquare of 

 the latitude of the firft degree:— That thtit Indian operations may reft 

 endrely on themfelves, I have adopted this method for computing a 

 fucceffion of ?im(3 degrees, beginning with the degree in latitude 

 9° 34' 44', which is 60472,^3 fathoms. The eigh'.h of thefe degrees 

 falls in latitude 16" 34' 44'', and*- is 605,09. 12 fathoms.— Now the degree 

 due tohdtude 16'^ 34' 42" as determined by the meafurement is 60512,78 

 fathoms, fo that there is; only a difference of 356 fathoms, a quantity 

 too inconliderable to affeft the elliptic hypothefis. — This is fuppofing 

 the, degree in latitude 9*^ 34 44: 'to be right, in which cafe the com- 

 preffion at the poles would be -^^y nearly. But if the compreffion -^l^-^, 

 as deduced from the general mean be fuppofed corred, and. the degree 

 in9^ 34' 44^''increafed to 60475,13 fathoms (fee Art. 16,) the next 

 degree in 10° Siii'y ^i^^ ^^ 60478,72, and thefe ufed will give 

 the compreffion jVt nearly : fo that by this method, the errors in ths 



