28 



fluctuations in the barometer. These disturbing causes will, 

 in any region, be much increased at the setting in of winter 

 and the commencement of hard frost ; for at this crisis a large 

 amount of latent heat will be liberated and will contribute its 

 influence to disturb the equilibrium of the air. A similar 

 crisis will occur at the end of winter on the breaking up of the 

 frost, and will necessarily be attended with similar results. 

 As the times of these crises appear to correspond in a remark- 

 able manner with the times of maximum disturbance of the 

 barometrical column, it seems but fair to infer that a relation 

 exists between the causes here stated to be in operation, at 

 the critical periods just named, and the periodical disturbance 

 of the mercury in the barometer indicated by Mr. Baxendell's 

 " dynamical curves." The correctness of this inference, or 

 the contrary, can only be established by future observation of 

 phenomena, and the collection of facts, many of them of a 

 kind seldom thought of hitherto as constituting elements for 

 the solution of problems in meteorology. 



