of the Barometer near Edinburgh. 



179 



It is well worthy of remark upon the preceding Table, that, by 

 means of the constants derived solely from the employment of ob- 

 servations in Latitudes 10 8" and 43 36', which have both a sign 

 of-}-, we obtain a result for Lat. 73 48' N. not merely giving a sign 

 of , but agreeing in amount with singular precision with that 

 which observation affords, proving that a negative oscillation, in- 

 stead of being in any way anomalous, is nothing else than we 

 might have inferred a priori for high latitudes. 



16. In order that the constants may express the general re- 

 sult of all these thirteen observations in the best possible man- 

 ner, we must reduce the sum of the squares of the errors to 

 a minimum. Let a + <*>' and + ' be the new values of the 

 constants upon this supposition; let also e be the error of the 

 above table, and E the reduced error. We then have 



and therefore, 



( + *') cos * 6 + (g + ') z= E, 



a' cos 2 



which gives the following thirteen equations of condition : 



