PROPERTIES OF SALINE SOLUTIONS — MACGREGOR. 78 



le^^el of the liquid is lower than at the outset. The amount of 

 the depression of the liquid in the tube is sometimes small, 

 depending apparently upon the amount of water which the 

 powdered caustic soda has already aljsorbed. The substance 

 shoukl not be too finel}^ powdered, as in that case it is likely 

 l)ot]i to have taken up a considerable quantity of water and to 

 carry down with it a considerable quantity of air. The experi- 

 ment requires only a few minutes to perform. 



(2.) The working hypothesis which I use when thinking of 

 the phenomena of solution, has led me to the conclusion that 

 elevation of the temperature of a solution will have, if not iden- 

 tically, at any rate in a general way, the same effect cm its 

 selective absorption of light, and therefore on its colour, as 

 increase in its concentration. All the experimental evidence 

 of which I can lind any record bears out this conclusion. 

 But, whether it holds generally or not, it may l»e shown, 

 1»y a striking lecture experiment, to hold in the case of 

 two salts, the chlorides of cobalt (CoCl,) and iron (P'eClg). 

 To do so, make a trough, for projection with a lantern, 

 having thin glass sides, aljout the size of a lantern-slide, the 

 glass sides being one or two mm. from one another. It may 

 readily be made by cutting a U-shaped piece from a sheet of 

 india rubber, and cementing the glass plates to its opposite sides. 

 Half fill the troufih with a saturated solution of either salt, and 

 fill up with a weak solution. If cobalt chloride have l)een used, 

 the solution in the lower part of the trough wall at ordinary 

 temperatures be of a purplish l:)lue, that in the upper part red j 

 and it will be obvious that increase of the concenti'ation of this 

 salt involves increase of blueness in the transmitted light. If, 

 now, a Bunsen flame Ije pla^'ed carefully over one side of the 

 trough, the solutions rapidly rise in temperature, and l:)oth are 

 seen to increase in Ijlueness, the saturated solution becoming deep 

 l:)Iue and the weak solution purplish red. If the iron chloride 

 have been used, the solution in the lower part of the trough, 

 l)efore heating, is seen to be of a deep orange color, that in the 

 upper part j^ellow ; and it is ol)vious that increase in the concen- 

 tration of this salt involves increase in redness. If, now, the 



