COLOUR OF WATER. 317 



these mountain rivers, and why should the 

 same element or substance assume such a 

 variety of tints ? 



HAL. I certainly have often thought 

 upon the subject, and have made some ob- 

 servations and one experiment in relation to 

 it. I will give you my opinion with pleasure, 

 and, as far as I know, they have not been 

 brought forward in any of the works on the 

 properties of water, or on its consideration 

 as a chemical element. The purest water with 

 which we are acquainted is undoubtedly that 

 which falls from the atmosphere. Having 

 touched air alone, it can contain nothing but 

 what it gains from the atmosphere, and it is 

 distilled without the chance of those impurities 

 which may exist in the vessels used in an 

 artificial operation. We cannot well examine 

 the water precipitated from the atmosphere, 

 as rain, without collecting it in vessels, and 

 all artificial contact gives more or less of 

 contamination; but in snow, melted by the 

 sunbeams, that has fallen on glaciers, them- 

 selves formed from frozen snow, water may 

 be regarded as in its state of greatest purity 



