ANALYSIS OF POTABLE WATE11S. 
65 
(13.) Lead. —This impurity may be derived from leaden cisterns or service pipes. It 
is easily detected by acidulating with two or three drops of acetic acid, and adding to 
the water about l-20th of its bulk of an aqueous solution of sulphuretted hydrogen. On 
looking down through a tube containing a column of water so prepared, 10 inches in 
length, upon a white surface, and comparing with it a similar tube tilled with the water 
without any addition, a brown discoloration, more or less intense, will be distinguished in 
the tube containing sulphuretted hydrogen, if lead be present; the amount of lead may 
be estimated by preparing a standard solution containing one ten-thousandth of its 
weight of lead, diluting the required quantity of this liquid with a known volume of 
distilled water, and comparing it with the water after the action of sulphuretted hy¬ 
drogen, in a manner similar to that directed to be followed in the colour estimation of 
ammonia (par. 10). The standard lead solution may be prepared by dissolving T831 
grain of normal crystallized lead-acetate in 10,000 of water, or OT831 gramme of the 
lead-salt in 1 litre of -water. 
(14.) Determination of Saline Constituents. —I have little that is new to add on this 
head respecting the actual quantitative estimation of the various constituents, but I 
should strongly recommend that in every case the amounts of the chlorine, sulphuric, 
nitric, and silicic acids, the lime, magnesia, ammonia, potash, and soda, as actually 
obtained by experiment, should be given, as well as the w r eight of the earthy carbonates 
separated during ebullition. If the potash and soda are not estimated separately, their 
joint weight should be determined by precipitating the chlorine or sulphuric acid from 
the alkaline chlorides or sulphates. By reporting the results of our experiments in this 
way, and in this way only, is a strict comparison of the analytical results of different 
observers possible without much unnecessary labour. 
On the representation of the exact form in which the different acids and bases are 
arranged in any water, chemists are not agreed, and it would be better to omit the 
practice of reporting the names of the salts altogether. If the plan thus recommended 
be adopted, the sum of the saline constituents would, of course, not correspond with the 
amount of fixed saline residue obtained on evaporation, but a general statement of the 
character of the water,—whether, for instance, calcareous, magnesian, or alkaline salts 
preponderated; whether carbonates, sulphates, or chlorides are most abundant,—would 
convey more real information to those who know little of chemistry than any enumeration 
of saline constituents, which no two chemists would report exactly in the same way. 
Of course, in all cases of importance, each determination shouldjbe made in duplicate; 
a decigallon or a half-litre being quantities "well suited to experimental work. 
It is almost needless to say, that in all accurate experiments the precipitates should 
be collected upon filters which have been previously soaked in dilute nitric acid (1 part 
acid to 30 of water), and afterwards thoroughly washed in distilled water and dried. 
It is obvious that the analysis must always be performed upon water absolutely freed 
from suspended matter, either by subsidence, which is best, or by filtration. 
The soluble silica is determined in the usual manner from the portion of water (02 
gallon, or 1 litre) evaporated down with carbonate of soda for fixed constituents. The 
ignited residue is treated with hydrochloric acid, adding distilled water, and then col¬ 
lecting the undissolved residue on a filter, igniting, and weighing. 
The filtrate must be treated with a slight excess of ammonia, immediately filtered, 
and the precipitate thus separated will contain the oxides of iron and manganese , the 
alumina , and the phosphates. They may be reported together. 
No direct determination of the quantity of carbonic acid which exists in combination 
with bases, can be conveniently practised, but the amount of mixed carbonates can be 
obtained with a fair approach to exactness by taking two Florence flasks—A and B—of 
similar size, counterpoising one against the other, and boiling down two decigallons of 
water in the flask A, concentrating to one-fourth of its bulk (or one litre boiled down to 
one-fourth), adding the water by degrees as it boils away. The ebullition must be 
conducted with the usual precaution of inclining the flask, so as to prevent loss by spirting. 
The concentrated water must then be filtered upon a small weighed filter, washing 
out the loose portions of the carbonate into the filter, drying the flask A and adhering 
carbonates at 212° (100 c C.), and sucking out the moist air, weighing when cold, coun¬ 
terpoising with the empty flask B, which should be placed alongside the flask A with 
the carbonates whilst it is cooling, the weighings being repeated as usual till after re¬ 
heating to 212° they become constant. The lime and magnesia in the precipitate can 
be estimated in the usual way if desired, and the deposited salts should be tested for 
