94 



Mr. W. W. J. Nicol on the 



of the variation in the weight of the contents of the Sprengel 

 tubes. This was found in three cases to vary from '0002 

 to '0004, that is '3 milligramme. A difference of 1 milli- 

 gramme in the weight of the solution corresponds to '00006 

 in the specific gravity ; consequently the limit of error is 

 •00002, and is generally less than that. No appreciable in- 

 crease of this quantity is caused by the fact that the weighings 

 were made in air, without correction ; for the Sprengel tubes 

 varied but slightly in size, and care was taken to use the same 

 tube throughout each series of experiments : these results then, 

 being comparative, are strictly comparable. The dilutions 

 were made by mixing equal volumes of salt-solution and water. 

 The quantities taken were ten times the specific gravity in 

 grammes and ten grammes of water. The error in volume 

 in no case exceeded '0003 cub. cent. The specific gravities 

 of the solid salts (finely powdered and sifted through muslin) 

 were determined in paraffin by means of a special form of 

 Sprengel tube, thus eliminating errors unavoidable when 

 bottles with movable stoppers are employed. The limit of 

 error is the same as that in determining the specific gravity 

 of the paraffin by means of the same tube. Eleven determi- 

 nations gave results varying from '81831 to '8184, with a 

 mean of '81836, error +'00005. The specific gravity of the 

 solid salt is therefore correct to the fourth place. 



Kremers, in 1852*, found that the specific gravity of a solid 

 salt was influenced by the temperature at which it was crystal- 

 lized from its solution ; and that if the solubility of the salt 

 increased with the temperature, the specific gravity of the solid 

 salt diminished with the rise of temperature of the solution 

 from which it separated ; if, however, the solubility diminished 

 with the temperature, the specific gravity increased. I have 

 repeated the experiments of Kremers, and the results are con- 

 tained in Table I. ; they agree substantially with his. There 



Table I. 



Salt. 



t°. 



Sp. gr. 



Salt. 



t°. 



Sp. gr. 



Salt. 



t°. 



| S P- g r - 



KN0 3 ... 



o 



20 

 110 



210355 

 209916 



Na 2 S0 4 



o 

 40 



110 



2-66180 

 2-66372 



NaOl 



o 



20 

 108 



2-16171 

 215494 



Difference ... -00439 + -00192 



- -00677 



Water at 20° =1. t°= temperature of crystallization. 





* Poggendorff's Annalen, vol. lxxxv. 



