392 



STATE BOARD OF AGRICULTURE. 



hydroxide added and the bridge readings taken. Each o])erator made 

 four sets of determinations on the same solution, the conditions being 

 altered each time so as to give dillVreiit readings for each set, thus 

 eliminating any error due to the oiKM-ator attemi)ting, unconsciously or 

 otherwise, to duplicate his ju'evious results. lOacIi one was required to 

 plot his own results and report indei>endently the amounts of ammonium 

 hydroxide re(piired to neutrali/.e the acid in Ihe quantity of citrate solu- 

 tion taken. The results are given in Ihe following table: 



TABLE VI. 



For the sake of comparison, averages of the several determinations, 

 made by each method, are given in Table VI. 



TABLE VII. 



The determinations by the corallin and litmus methods were made 

 on the same solution but the determinations by the conductivity method 

 were made upon a solution of slightly different strength. The advan- 

 tages of the conductivity method over the other Iwo is well illustrated 

 in the above table. The results by the corallin and litmus methods are 

 very variable, the greatest variation in the first case being 'A.&2 while 

 in the second ca.se it i.s 1.70. 



The results by the conductivity method are very much closer, the 

 maximum variation in the averages l)eing only 0.43. If we assume 

 the average of all the series to l>e the true result, then the maximum 



