202 DR. R. ANGUS SMITH 



This shows that^ for the precipitate on the plan of Tables 

 I. and III., lime-water is more than necessarily constant, 

 and even enough for Table V., and may be used in some 

 cases to save the purchase of a balance. 



Some years ago I proposed rosolic acid as an agent for 

 ascertaining neutrality. It shows the neutral point with 

 extreme sharpness when oxalic or any liquid acid is used ; 

 and we can tell to a small drop when to cease pouring. 

 If all caustic solutions were coloured with it, it would help 

 to point them out, as well as serve instead of litmus or 

 turmeric. I did not obtain such successful experiments as 

 I could have wished when trying it with carbonic acid 

 lately. The last traces of colour are difficult to remove ; 

 I hope to bring it on further some day. It is curious that 

 lime- or baryta- water take up carbonic acid much less 

 readily when rosolic acid is present. The latter portion of 

 the experiment is aflFected chiefly when the solutions be- 

 come weak. It is as if the resinous consistence of the acid 

 repelled the gas. When a liquid is used, such as oxalic acid 

 in solution, the rapidity of action is very great, the same 

 resinous quality causing the rosolic acid to shrink into its 

 shapeless and colourless state with great suddenness. 



It may, however, be remarked that weak solutions of 

 baryta and lime take up carbonic acid very slowly of them- 

 selves, although rosolic acid renders the absorption still 

 slower. 



Manganates and Ferrates as Tests for Carbonic Acid. 



In a former paper I mentioned that the carbonic acid of 

 the air was capable of being estimated by observing its 

 action on manganates. The green manganate was prepared 

 in the usual way, or by adding caustic alkali to the so-called 

 permanganates, which, however, from this action appear 

 rather as bi-manganates. The amount of acid required to 

 convert the green to red may be found readily by a solu- 



