§8 Mr. M. Carey Lea on the 



■ Note added June 5, 1892. 



In further tests with this photometer the mean errors have 

 generally fallen "below 1 per cent. 



At first I tried other angles than 45°. For whitewashed 

 metal screens this angle appears to be suitable. I find, how- 

 ever, that it is preferable with paper screens to use an angle 

 of incidence of 35°, and I understand that this angle is used 

 in the Thompson- Starling photometer. 



When the light to be measured is subject to small varia- 

 tions, the direction and the magnitude of the variations are 

 plainly visible by the movement of the band of uniform tone. 

 This photometer has been chiefly used for such measurements, 

 and seems to be well adapted for them. 



VII. On the Nature of certain Solutions, and on a New 

 Means of Investigating them. By M. Carey Lea*\ 



THE three strong acids with which we are best acquainted 

 have this in common, that they all form two classes of 

 compounds, — the one perfectly neutral and perfectly stable in 

 solution, the other class instantly decomposing when it is 

 attempted to dissolve them in water. As types of the first 

 class maybe taken the alkaline salts. Of the second, mercuric 

 sulphate, bismuth nitrate, and stannous chloride offer examples, 

 as also the thallic salts of all three. 



Between these limiting cases are certain intermediate ones 

 as to which our knowledge of the form which the salts take 

 in solution is not very exact. It need scarcely be men- 

 tioned that what has been said does not refer to dissociation 

 into anions and kathions, but into separation of free acid and 

 basic salt, an entirely different question. 



In the case of sulphates — -to a consideration of which this 

 paper will be devoted — we know that the number of salts of 

 the metals which give a purely neutral reaction with litmus 

 is comparatively small ; that all the normal sulphates of the 

 heavy metals show an acid reaction even after any number of re- 

 crystallizations, although these may be made from solutions 

 rendered alkaline by the addition of free alkali, and that all the 

 normal sesquisulphates have an acid reaction. The question 

 is as to the meaning of this acid reaction, and whether in any 

 or all cases free acid is present. 



The uncertainty that exists depends partly upon the im- 

 perfect nature of our indicators. Litmus, which is perhaps 



* Communicated bv the Author. 



