ON THE SENSITIVENESS OF INDICATORS. 



275 



cither constituent, we find that the flat part of the curve is much narrower 

 and much less flat. Instead of there being a number of indicators which 

 would give accurate results, we find only methyl red in one case, and 

 phenolphthalein in the other. Methyl red gives not only an accurate but 

 a sharp end-point in ammonia titrations ; phenolphthalein, instead of also 

 giving an accurate result, gives an extremely inaccurate one, certainly not 

 within 2 or 3 per cent. Methyl orange gives a better result in this cate 

 than phenolphthalein, but still not a sharp end-point. These facts have 

 of course been long known, but the curves show well the magnitude of 

 the errors involved — an important factor. In the titration of acetic acid 

 by soda, methyl red gives an inaccurate and extremely bad end-point, 

 phenolphthalein a sharp and accurate one. If the base or acid used is still 

 weaker, it becomes very difficult to find a good indicator, and finally im- 

 possible ; the flat part of the curve in fact tends to disappear until at last it 

 does so altogether. As an example we can take aniline. Here, owing to 



51 



50 



_J 



oMngE 





N/ 



m 



100 



h 



% 



tu 



49i 



r 



100 



Jid 



ANILINE 



ten 



QQJ 



CHfflOH 



by /oo N HO H 



io*" io*~ io" 5 io* itFlo*" ib" 9 



Fig. 3. 



the great hydrolysis of aniline hydrochloride in solution, the concentration 

 of hydrions at the ' equivalent ' point is as high as 10~ 3 ' 5 , and a large excess 

 of acid or base produces only a slight change in this value. Since such 

 concentrations of H° ions do actually affect the colour of methyl orange 

 appreciably, it might be possible to titrate aniline to about one per cent, by 

 using a comparison solution and a colorimeter, but evidently it is quite 

 impossible to find an indicator capable of giving even moderately accurate 

 results without this means. 6 



A curve showing the titration of ammonia by acetic acid is also given : 

 it will be seen that even though the concentration of hydrions at the 



8 The concentration of hydrogen ions at the (true) end-point of a titration 

 alters of course with the temperature, being in general about three times as small at 

 0° as it is at 25°. Reference to the curves given will show that this difference has 

 very little effect on the probable accuracy of the titration, except when the acid or 

 base titrated is extremely weak. It must be remembered that the effect is consider- 

 ably lessened by a simultaneous decrease in the dissociation constant of the indicator. 



T 2 



