16^ M, Uamond^s Instructions for the Application of [SfiPT. 



discrimination, and experience employed ; that we do not learn 

 all this as we learn to read, to calculate, or to translate the terms 

 of a formula. Let us read then, and reflect upon the works of 

 De Luc, Saussure, the memoirs of Pictet, the books of philoso- 

 phers who have devoted themselves to the scientific employ- 

 ment of meteorological instruments ; and above all, let us con- 

 tinue our own observations for a long time before we trust to 

 them, and a still lonoer time before we yield to the temptation 

 of drawing inferences from them ; for the goodness of observa- 

 tions depends on a multitude of minute precautions which habit 

 alone renders constantly present to the memory ; and the vali- 

 dity of the inferences rests on considerations too numerous to be 

 at the disposal of a mind not rendered familiar to them by time 

 and practice. 



Sedentary Observations * for determining the Mean Pressure at a 

 particular Place. 



I commence with considering the method of conducting 

 observations of the stationary kind for determining the mean 

 pressure of the atmosphere at any place ; for these form one of 

 the principal foundations on which the science of meteorology 

 rests, and above all because they are the means of affording 

 marked periods for the operations of levelling. 



Sedentary observations have a very hmited utility if they are 

 not comparable ; that is to say, if equal pressures are not con- 

 stantly expressed by equal heights of the mercurial column. 

 We may readily conceive that this fundamental condition is but 

 ill fulfilled by the rude instruments with which the generality of 

 observers content themselves, and often not even by the more 

 costly ones with which the cabinets of the curious are adorned. 

 '•>>■ It must be presumed then that the observer has a barometer 

 -carefully constructed : the mercury perfectly pure and well 

 boiled ; the tube as free from air and moisture as it can be ren- 

 dered : the scale leaving nothing to be desired on the score of 

 accuracy ; he beheves himself to possess a very good baronieter. 

 It is doubtless so far good ; but this is not enough ; and the 

 chief condition is not fulfilled unless we are assured that it is 

 €omparahle. What is the interior diameter of its tube ? Wha^ 

 means have been taken to correct the effects of capillarity ? Is 

 it of the siphon or cistern construction '/ What precautions have 

 been taken to secure a constant level, and an invariable point of 

 departure ? Such are some few of the questions which must be 

 satisfactorily answered before the instrument can be allowed to 

 inspire confidence. 



Let us, however, suppose all these conditions fulfilled, still it 

 must be further asked, if the barometer be comparable, are the 

 observations equally so ? At what hours and under what circum- 



♦ " Observations SCslentaires.'* 



