of a J^ew Hygrometer. 105 



a thermometer placed in the shade, and another suspended 

 in an ordinary room, free from the sun's direct rays, there 

 are from four to five degrees difference. 



When Professor Daniell's Hygrometer is employed in 

 Madeira, the temperature of the air ought to be taken, not 

 from the thermometer attached to the instrument, but from 

 another thermometer, protected as much as possible from 

 every source of radiant or reflected light and heat. 



I think it right to state these facts, in order to inform 

 the inexperienced operator, that even the best instruments 

 will give erroneous data, provided every circumstance be 

 not considered when making observations with the same 

 instrument in different climates, for the only sure method 

 by which one climate can be compared with another, is to 

 make the experiments under precisely similar circum- 

 stances. 



In my Hygrometer, as in all Dew-point Hygrometers, it 

 is absolutely essential to observe great accuracy in taking 

 the temperature of the shade ; because an incorrect indica- 

 tion of the shade will necessarily lead to erroneous results 

 in obtaining the humidity by both instruments. 



It is, however, to be remarked that when both descrip- 

 tions of Hygrometers are employed, the one will correct 

 the other; the Dew-point Hygrometer giving greater dry- 

 ness, and mine giving greater humidity than really exists 

 in the atmosphere, each in equal ratio : so that the mean 

 between the results of both instruments will indicate the 

 actual condition of the atmosphere. 



It is important to know this fact when exact results are 

 required, and there is difficulty in obtaining with precision 

 the true temperature of the shade, as often happens in 

 southern latitudes. 



If the degree of dryness on the thermometric scale be alone 

 required no inaccuracy can result in my Hygrometer, even 

 when the temperature of the shade is incorrectly taken, 

 because the temperature does not affect its indication in this 

 respect ; whereas the Dew-point Hygrometer cannot possibly 

 possess the same advantages, because if an error is com- 

 mitted in ascertaining the temperature of the shade, it will 

 give a dryness exceeding the real condition of the air by 



