27 
filtrate. It is to be regretted that the filtering could 
not have been done through a Gooch filter, as then the 
precipitated silver haloid could have been collected and 
weighed, and so would have formed an additional check 
on the values obtained by titrating the filtrate. 
This filtrate was poured into a porcelain dish, a 
drop of iron alum solution added, and titrated drop by 
drop with the sulpho-cyanide solution till a red colour 
appeared. To this was then added enough silver 
nitrate solution to just destroy the colour, and this 
process of alternately titrating with the sulpho-cyanide 
and the silver solution was frequently repeated in a 
zigzag manner till a great number of determinations 
of the end point were obtained, the mean of which was 
considered accurate. 
(c.—Usine Porasstum CHromatE as INDICATOR. 
As in the previous case a sample of about 10cce. of 
the water was first titrated to find about how much 
silver solution was required. 
In the accurate determinations 10cce. of the sea water 
were mixed with some distilled water in a beaker, and 
then about O’dce. less silver solution than was required 
for complete precipitation was added. The mixture 
was well mixed and a drop of the chromate solution 
added, and then the silver solution was run in drop by 
drop till a permanent change of colour was obtained. 
The end point was obtained repeatedly by zigzag 
titrations with the salt and silver nitrate solutions. 
A mean of the values obtained was taken as correct. 
Considerable errors oceur in both methods, but more 
especially in this, owing to the difficulty of obtaining a 
standard of colour which represents the end of the 
reaction. 
