Self-recorded Breaks in the Properties of Solutions. 473 



plotted against time. The dotted lines formed by the nume- 

 rous experimental points in the upper drawing have been 

 extended a little by continuous lines in order to make the 



Fi-. i. 



changes more easily discernible. To ascertain the strength 

 at which any change in the upper drawing has occurred, it 

 is only necessary to take the reading of the lower curve for 

 the same time. The temperature-readings form five lines 

 which within experimental error are absolutely straight, and 

 owing to the well-marked angles at which these lines meet, 

 and to the large number of experimental points on them (as 

 many as 55 in one case) , there can be no doubt as to their 

 real nature. Of the four changes which this figure exhibits 

 two are incontestably breaks : A is the point at which the 

 stream of acid was started, and D that at which it was 

 stopped ; yet these are no better marked than those at B 

 and C, which, therefore, must equally be admitted to be 

 breaks. 



Looking at the changes near A and D it will be seen that 

 these occur about one minute after the stream of acid was 

 started and stopped respectively ; this represents the time 

 taken by the acid to get mixed with the water, and the lag of 

 the thermometer behind the temperature of the liquid. 



The break at C was not so well marked as it might have 

 been had the experiment been continued for a longer time. 

 Another experiment illustrated in fig. 2 shows it better. 

 Here, however, the stream of acid being faster, there were 

 not sufficient observations with the weakest liquid to show 



