693 



uninterrupted ascending and descending move^ 

 ment. At the periods when the mercury during 

 twenty-four hours, attains the maximum, and 

 the minimum, m, n, m and ri, the direction of 

 the movement remains constantly the same, 

 from m to n, and from m to ri, whatever may 

 be the hours in different places of the earth, to 

 which the concave, or convex summits of the 

 curve of diurnal variations correspond. We 

 scarcely find in thousands of American obser- 

 vations, one or two exceptions to the laws I 

 have ascertained. Accustomed to an uninter- 

 rupted regularity, the observer is so much 

 struck by the slightest anomaly, that he is often 

 tempted to attribute it to some negligence in 

 the observation, or the want of perpendicularity 

 in the instrument *. At Cum ana, for instance, 

 on account of this continuity of the move- 

 ments, one day and one night suffice to ascer- 

 tain the type of the progress of the barometer ; 

 while in Europe, we must take the mean, not 

 of a decade, but (as we shall soon shew), at 

 least of twenty or thirty days. 



II. Epochas of the maxima, and the minima. 

 Duration of the stationary state. There is 

 something vague in the manner of indicating- 



* See above, in the observations at Cumana, August 24th 

 and 30th (Vol. vi, p. 666). 



