ON SALTS* 241 



Iron of the falts, fince It caufes thedcnfity of the air to vary; 

 confequently there (hould be an agreement between the vari- 

 ations of the barometer and (he deliquefcence of falts. 



The variations of temperatures, by dilating, or by conden- — thermomctri- 



fing the mafs of the atmofphere, fhould alfo occafion changes J al &<>*]& alfo 

 -7, . , , r ...... , i - i have their influ- 



in the proportion ot water abforbed by the ialts, on which ac- ence# 



count it would be ufeful to obferve the thermometer. 



I thought moreover that one fait had not only more or lefs Deliquefcent 



attra&ion for the water contained in the air than another, but falt i fhou ' d a r" 

 .i i- ^- - . ,., •/--,/- r • • • tract moft when, 



that this attraction varied likewile in the lame talt in proportion i ea ft faturated. 



as it had loft or abforbed water. I hoped by thus comparing 

 the deliquefcence and efflorefcence of falts with the ftate of 

 the different meteorological inftruments, to obtain refults fuffi- 

 ciently conftant to eftablifh a theory of deliquefcence or ef- 

 florefcence. I hoped alfo to be able to ufe the falts themfelves 

 as inftruments of meteorological obfervation ; but experience Experience did. 

 proved that reafoning apparently founded on the trueft theory ^/""ofSon 

 frequently deceives expe&ation. It is neverthelefs neceffary 

 to attend to negative facls, which fometimes areas ferviceable 

 to fcience as thole of a pofitive nature. 



I did not find a tingle fait which feemed to have the Ieaft None of the falts 



conformity with the ftate of the barometer, hygrometer, or^'f^.V 1 ' 11 

 J _. ■ '■ . J ° or lofe weight in 



thermometer. Gn the lame day many falts increafed confider- conformity to 



ably in weight, while others indicated a flow progrefs. Some 11 ^ 01 " 10 



had but a fmall attraction, when the hygrometer {hewed a 



great degree of humidity, and were moft deliquefcent when 



the air feemed moft dry. Atmofpheric preffure never had the 



Ieaft agreement with the increafe of weight of a fait, and the 



thermometer having varied but half a degree during the courfe 



of the experiments, does not furniih any obfervation on the 



influence of temperature. It is therefore impoffible to explain 



by the meteorological changes any of the variations which I 



obferved in the deliquefcence, or the efflorefcence of falts. 



Efflorefccnt Salts. 

 I weighed exa&Iy 288 grains of fulphate of foda, of phof- Experiments. 

 phate of foda, aud of carbonate of foda, which three falts are fLts^phlte, 

 confidered as the moft efflorefcent, and placed them in a dry phofphate and 

 and airy fituation, after having carefully dried the capfules ^ r p b ^ e of foda 

 which contained them. I put alfo in the fame place an hygro- 

 meter, a barometer, and a thermometer : the three ialts fhewed 

 the following refults. 

 Vol, XII.— December, 1805. R Sulphate 



