ANALYSIS OF POTABLE WATERS. 
59 
I believe that much -would he gained, both in precision and in the comparability of 
our results, by performing every detail in exactly the same way, time after time, as in 
the common methods of assaying. In such cases it is astonishing how closely the re¬ 
sults obtained coincide with each other when the identical alloy is repeatedly subjected 
to the process by the operator, who may be, nevertheless, quite unconscious of the strict¬ 
ness of the ordeal which he is undergoing at the hands of the client who is testing the 
accuracy of his work. In a resume of the nature of that which I am proposing to make 
on the present occasion, any great amount of novelty will not be expected. I have se¬ 
lected trom every available source with which I am acquainted methods which, if not 
minutely accurate in the scientific sense of the word, experience has proved to be at 
least practically correct, and for the most part easy and rapid in performance.* 
(2.) Mode of observing Colour of a Water. —After examining the appearance of the 
water in a beaker, a portion of it should be placed, as practised by Dr. Letheby, in a 
tube two feet in length and an inch or more in diameter, closed at the bottom with a 
flat plate of glass. A similar tube is filled with distilled water, and compared with the 
water under trial by placing the two tubes side by side, and looking down upon a white 
plate. The water usually has a shade of yellow, green, or brown, if organic impurity 
be present, and the amount of colour can then be estimated and may be noted in terms 
of an arbitrary scale.f 
(3.) The amount of sediment held in suspension in a gallon or half a gallon, or say 
three or five litres, is determined by allowing the water to settle, drawing off the clear 
water with a siphon, collecting the sediment on a weighed filter, and drying at 284° 
(140 C.); and the loss by ignition is afterwards ascertained when the filter and its con¬ 
tents are burned, and the residue is weighed. 
(4.) The taste of a water, as was pointed out by Dr. T. Clark, should be determined 
both at the ordinary temperature and when it has been warmed to between 86° and 95° 
(30" and 35° C.), at which point any unpleasant flavour is much more distinctly per¬ 
ceptible than when the water is cool. 
(5.) The odour will also be developed under similar circumstances, and will become 
more marked if a little lime or baryta-water be added before warming the water. It 
may be best observed by placing a few ounces of the water in a flask partially filled, and 
drawing air through a tube from the upper part of the flask into one nostril, whilst the 
other is kept closed. 
(0.) The hardness of a water, as determined by the soap-test of Dr. Clark, furnishes, 
of course, one of the most important data in relation to the fitness of such a water for 
domestic purposes; but all the particulars relating to the precautions needed to attain 
accuracy in this operation upon a water before it has been boiled have been so thoroughly 
worked out by Dr. T. Clark and Mr. D. Campbell, that I have nothing further to add. 
I may remark that in taking the hardness of a boiled water , a steady ebullition of not 
less than one hour should be maintained before determining the extent to which soften¬ 
ing has been effected. The flask, of course, must contain a weighed quantity of water, 
and after the boiling the amount must be made up to the original weight by adding 
pure distilled water. A simple condenser, consisting of a tube about a third of an inch 
in diameter, four feet long, cut off obliquely below, and resting, by a bulb blown near 
its lower extremity, upon the neck of the flask, will reduce the loss by evaporation to ail 
insignificant amount. 
(7.) Evaporation and Incineration. —The most difficult, and at the same time most 
important problem in the analysis of a water is the determination of the amount of the 
organic matter , as it is loosely termed, which it holds in solution. The plan, often prac¬ 
tised, of simply evaporating down to dryness, weighing, and then igniting and weighing 
a second time, always leads to inexact results, and commonly produces an over-estimate 
of the amount of organic matter, owing to the loss especially of hydrochloric acid and car¬ 
bonic anhydride, if much magnesian salts are present. 
It is better to add a known quantity of fused carbonate of sodaj to the vessel 
* Since this paper was read, one or two alterations have been made in detail, in conse¬ 
quence of the remarks of friends, either at the meeting of the Chemical Society or in conversa¬ 
tions since. 
f If iron be present, special attention must be paid to this impurity. 
T In this and some other cases we have changed the names, and omitted or changed the 
F 2 
