302 Daniell on a new Hi/gro>neter. 



librium, are very perceptible. The bulb of the thermometer 

 d. e. is not quite immersed in the ether, that the line of greatest 

 cold may pass through it. The greatest difference that I 

 have observed in the coarse of four months' daily experiments 

 between the external thermometer k. I. and the internal one e. d. 

 at the moment of precipitation in the natural state of the 

 atmosphere, was 20 degrees. In very damp weather the ether 

 should be slowly dropped upon the ball, otherwise the de- 

 scent of the thermometer is so rapid as to render it impossible 

 to be certain of the degree. In dry weather on the contrary, 

 the ball requires to be well wetted more than once, to produce 

 the requisite degree of cold. It is almost superfluous to 

 observe, that care should be taken not to permit the breath to 

 affect the glass. With these precautions the observation is 

 simple, easy and certain. 



When the instrument is required to act merely as a weather- 

 glass, to predict the greater or less probability of rain, Sfc, 

 which is the commonest use to which it can be applied, the 

 difference between the constituent temperature of the vapour, 

 and that of the air, is all that is necessary to be known. The 

 probability of rain or other precipitation of moisture from the 

 atmosphere, is in inverse proportion to this difference. The 

 journal of observations annexed will prove that for this pur- 

 pose it is more to be depended upon than any instrument that 

 has yet been proposed. For example, upon reference to the 

 Table it will be found, that on the 29lh of August the baro- 

 meter fell continually all. the day, as it had done all the day 

 before. The hygrometer, however, shewed the great elasticity 

 of the vapour, and its little inclination to deposition. The 

 difference of the temperatures varied from 13° to 17°, and 

 there was no rain. The next day the barometer continued to 

 fall. In the morning, which was fine, the difference had fallen 

 to 10°, and in the afternoon to 2°. Heavy showers ensued, 

 which were not however of long continuance. On the 3d of 

 September, the barometer rising the whole day, the hygro- 

 meter denoted rain, which came down in heavy showers in the 

 evening. On the 29th and 30th of the same month, the 



