1818.] of the Profile of Mount Jura. 367 



zenith distances, observed by skilful mathematicians furnished 

 with repeating circles, he finds sometimes discordance between 

 the two methods, and great probability in favour of the first ; 

 sometimes agreement ; and in that case a prodigious advantage 

 in facility and dispatch on the side of the barometer ; " and the 

 accuracy of the trigonometrical method," he observes, " has 

 likewise its bounds. When we compare together the total and 

 partial measures which it has furnished us, we are forced to 

 acknowledge that none can be depended on nearer than one or 

 two metres. This uncertainty indeed is but small ; but that 

 which attends barometrical measurements is not greater ; and if 

 we employ separately the different series which have concurred 

 to the determination of the mean angle, these exhibit disagree- 

 ments which far exceed the limits within which the errors of the 

 barometer are confined." 



" The imperfection of the two methods comes from a common 

 source ; and the two instruments are equally defective when the 

 disposition of the atmospherical strata is such that the gradual 

 decrement of heat and humidity is altered or inverted. But this 

 disorder exhibiting itself to the trigonometrical instruments by 

 anomalies in the refraction, appears to exercise on the results an 

 incomparably greater effect than when it alters the ratio of the 

 pressures indicated by the two corresponding barometers. In 

 this last case, a small number of observations is sufficient to 

 compensate for the errors ; but in the first case a very consider- 

 able number are requisite : and M. Broussaud and myself have 

 convinced ourselves that barometrical measurements taken with 

 care are preferable to trigonometrical measurements themselves, 

 whenever these last do not repose on reciprocal and veiy nume- 

 rous observations, made with excellent instruments and by very 

 skilful persons, with all the conveniences and all the time 

 requisite for such operations." 



Explanation of the Profile of Jura in Plate LXXXVI. Places 

 denoted by the Letters. 



a, Geneva. 



b, Gex. 



c, Mount Colombier. 



d, La Faucille. 



e, Valley of Mijoux. 



f, Culminating point ; vallon tourbeaux. 



f, Sept-Moncel. 

 , Highest point of the road. 

 /', St. Claude. 

 k, Bienne (river). 

 /, Chateau de Prax. 

 m, Roche d'Antre ; signal. 

 u, Moyr.iiH. 



