Duniell on a new Hi/gromcter. 325 



the Hinnilaya measurements, is a very rough calculation, and 

 the amount of the error is probably below the truth ; it will 

 however sufficiently demonstrate the nature and importance of 

 the correction. In higher latitudes, and in the winter season, 

 the state of the vapour may more safely be disregarded, for its 

 pressure increases much more rapidly in the higher part of the 

 tiiermometric scale for every degree of temperature, than it 

 does in the lower ; its influence, however, under all circum- 

 stanees, is sufficiently great to make it of consequence, when any 

 thing like accuracy is required. 



I shall now conclude this paper. No one can be more aware 

 than myself of the incomplete state in which I have presumed 

 to bring forward these observations. More time, and better 

 opportunities, were required to attain that accuracy which is so 

 desirable in experimental inquiries. London, moreover, is perhaps 

 the worst place in the world for meteorological pursuits. Observa- 

 tions upon clouds, dew, and winds are almost precluded, 

 and any comparison of heights is very limited indeed. Know- 

 ing, however, that the great value of the instrument which I have 

 contrived, must be derived from the number, extent, and com- 

 parison of the experiments to be performed with it, by different 

 observers, in different situations, I have thought it more for the 

 advantage of science to bring it forward at once, trusting to the 

 candour of the learned to allow of the validity of the excuse. In 

 the meantime, I shall continue my observations, and reserve 

 what further I have to say for a future paper in the Journal, when 

 time shall have increased their number, and confirmed their 

 conclusions. 



The instruments are accurately constructed, and packed in a 

 box for the pocket, by Mr. Newman, Lisle-Street. 



