Mr. Daniell on the Barometer. 89 



of their oscillations : but I have not discovered any which have 

 been continued for a sufficient length of time, with the same in- 

 struments, to answer this purpose satisfactorily. Instances abound 

 of observers having taken the pains to re-boil their barometers 

 from air having obtained admission, in some unknown way 

 which has always been attributed to accident ; but the fact of 

 their gradual deterioration cannot, in this way, be established so 

 completely as might have been supposed. 



The register of the Royal Observatory at Paris has only been 

 published since the year 181G, in the Annates de Chimie ; a pe- 

 riod which is not sufficient so far to neutralize the annual oscil- 

 lation as to afford the means of a satisfactory comparison. 

 Mr. Howard, however, in his work upon the climate of London, 

 states the mean height of the Royal Society's barometer for ten 

 years, from 1797 to 1S06, to be 29.882 inches, while for the ten 

 succeeding years it is only 29.849 inches, which gives a depres- 

 sion of .033 inches in that interval. 



The observations of the following ten years will not, I fear, 

 be available in the same comparison, from the carelessness with 

 which they have been made. 



The difference in the height of the old and new barometer, 

 which have now been placed side by side, was, in the latter part 

 of the year 1824, .07 inch, upon a mean of twenty observa- 

 tions ; the new barometer standing upon the average so much 

 higher than the old one. Whether this be wholly owing to de- 

 terioration, it is not possible to say ; for the old barometer does 

 not appear to have been boiled : but from the well known accu- 

 racy of Mr. Cavendish, under whose superintendence it was con- 

 structed, it is impossible not to ascribe a great portion of it to 

 this cause. The mercury of this instrument is now thickly 

 studded with air-bubbles of much larger size than those of the 

 new barometer ; and when I last examined it, some of them 

 were just upon the point of making their escape. 



I now feel that I ought not to allow this opportunity to pass 

 without making a few observations upon the new meteorological 

 register lately published in the Philosophical Transactions. I am 

 aware that, in so doing, I shall run the risk of being again de- 



