THE ISOCHLORS OF NOVA SCOTIA. VICKERY. 361 



either in the laboratory of Bloomfield lU^h School, or in 

 a small laboratory fitted up for the purpose at the writer's 

 home, but in all cases the utmost precautions were observed 

 as regards lighting and to guard against any chlorine intro- 

 duced by improperly cleansed apparatus. The same silver 

 nitrate solution which was prepared in the laboratory at 

 Dalhousie University was used throughout the work, and 

 was kept carefully protected from light and pollution. 



No data was accepted without being carefully checked 

 by two, and in most cases three, separate determinations 

 of each sample. Wherever possible a number of samples 

 were collected from a small area, and the lowest of these 

 taken as the normal for that vicinity. The burette used 

 in the titration was one containing 10 cc. of the solution, 

 ringed around, and read by means of a lens. Great care 

 was exercised in keeping this burette in perfect condition 

 during the analytical work. The use of a graduate in measur- 

 ing out the samples for analysis was necessitated, as above 

 mentioned, by the inconvenience of carrying about a 100 cc. 

 pipette, and it is believed that the error in reading this is 

 well below the experimental error in reading the burette. 

 In draining this graduate into the porcelain casserole a 

 definite time interval was always allowed to elapse, and as 

 far as possible all observations were made under identical 

 conditions, so far as light was concerned, so that the experi- 

 mental results at least would be strictly comparable. 



The determination of the exact endpoint in white porcelain 

 dishes was found to be somewhat easier than in the ordinarily 

 used 100 cc. Nessler Jars. When dealing with water dis- 

 coloured wdth peat, use was at times made of a drop or two 

 of the chloride solution, to make sure that the endpoint had 

 been attained, although, as mentioned above, in every case 

 the standard of comparison was an identically prepared 

 sample, and in cases where the chlorine content was high, 

 and the cloud of precipitated chlorine rendered comparison 



