92 CALIFORNIA ACADEMY OF SCIENCES. 



The titration of the 25 c.c. obtained from the 250 c.c. of 

 sea- water is done in exactly the same manner as that of the 

 normal solution of potassium bromide, except that the ob- 

 servation of the first drop from the burette requires special 

 attention. In the presence of iron (from the zinc) or cop- 

 per, both being under the circumstances ferrous or cuprous 

 salts, no bromine will be set free until they have been con- 

 verted by the chlorine into ferric and cupric salts. There- 

 fore, before commencing on the fractional titration, I empty 

 the 25 c.c. into the flask, add one cubic centimeter chloro- 

 form, also one drop from the burette, shake, and observe 

 whether the chloroform shows coloring. If it does, I 

 evaporate the drop of chloroform by boiling the contents of 

 the flask while keeping them in agitation, cool, return the 

 solution to the graduated cylinder, fill up to 25 c.c. and 

 titrate in the manner described above, adding the one drop 

 from the burette used for coloring the chloroform to the 

 account. 



The more experienced analyst will, however, be able to 

 recognize the first color without requiring chloroform, and 

 thus avoid boiling the solution previous to titration, which 

 may cause a slight loss by evaporating some hydrobromic 

 acid, unless sodium carbonate is added. For the titration, 

 the solution has then to be made acid again. Whilst by 

 this preliminary test traces of iron, manganese, copper and 

 sulphurous gas can be detected and made harmless, the 

 presence of hyj)osulphite of zinc (formed by the action of 

 sulphurous gas on zinc) affects the assay more seriously, as 

 it is less readily oxydized, completely only when the titrated 

 solution is boiled to evaporate the free bromine. This 

 causes formation of hydrobromic acid — hence the necessity 

 of guarding against the presence of sulphurous gas before 

 zinc is added. Every analyst will judge what other sub- 

 stances may be present in his solution which might inter- 

 fere, and how to remove them. 



It remains to state that the titration as described may also 



