David Sommerville.

Practical sanitary science : a handbook for the public health laboratory online

. (page 4 of 27)
Online LibraryDavid SommervillePractical sanitary science : a handbook for the public health laboratory → online text (page 4 of 27)
Font size
QR-code for this ebook

Stannic salts in acidified solution :

1. HjS: yellow precipitate of SnS.2, soluble in both yellow and
colourless ammonium sulphide; soluble in KOH on heating; re-
precipitated b}^ HCl as yellow SnSg from both solutions.

2. HgClg: no precipitate.

3. AuClg : no precipitate.

Quantitative Estimation of Tin [Stannous or Stannic).- — In a
measured quantity of water, concentrated if necessary to a small




bulk, precipitate the Sn as sulphide. Stand in a warm place till the
smell of H.,S has nearly- disappeared. Filter. Wash well. Dry.
Ignite in the air into SnO.,. Weigh, and calculate the Sn. [Inciner-
ate the filter-paper apart from the precipitate, and add tlie ash to
the crucible containing the SnOo-]

No tin should be present in drinking water.

Iron. — Ferruginous waters are found in mountain limestone,
chalk, Bagshot sands, and greensands. They are generally opal-
escent, and slightly 3'ellow in colour. The metal occurs as a
bicarbonate which is readily converted into an insoluble carbonate,
and also oxidized into the w^ell-known ' rust ' — hydrated ferric
oxide, FcaO^, H-iO. Ferrous salts decompose nitrates, absorbing
and producing nitrites, which in turn are further reduced to
XH3. This reducing action accounts for the free NH3 often found
in pure waters derived from the greensands and other strata.

Chalybeate waters may be quite clear when drawn, but as
oxidation of the Fe proceeds they become turbid and more or less
brown. The insoluble and highly oxidized particles dissolve on
the addition of a little dilute acid. Such turbidity may have its
origin in iron pipes, cisterns, etc., in addition to strata.

Two classes of iron salts exist : ferrous, in which Fe is divalent, and
ferric, in which it is trivalent. They may be readily distinguished
b}' the three reagents, potassium ferrocyanide, K4Fe(CN)g, potassium
ferricyanide, K3Fe(CN)fi (a solution always being made from the crys-
tals immediately before use), and potassium sulpho-cyanide, KCNS.


Ferrous Compound.

Ferric Compound.

K4Fe(CN)6 -

K3Fe(CN)6 -

Light blue precipitate, be-
coming dark blue on oxi-
dation by the air, HNO3,
or Br.

Dark blue precipitate;
Turnbull's blue insoluble
in HCl.

No red colour.

Dark Prussian blue, in-
soluble in HCl; turned
brown by KOH.

No precipitate.

Blood-red colour (no ])re-
cipitate). Colour de-
stroyed by dropping a
few drops into a solution
of HgClg.


Sulphuretted hydrogen passed through a solution of a ferric salt
reduces it to the ferrous state, with deposition of S. H^S gives no
precipitate with a ferrous salt in acid solution.

(NH4)2S precipitates from a ferrous salt black ferrous sulphide.
This reagent reduces a ferric salt to the ferrous state, and then
precipitates ferrous sulphide with S.

Except in connection with greensands, ferrous iron is rarely, met
with in water work, ferric salts alone being found.

(NHJaOH produces with ferric salts a reddish-brown, flocculent
precipitant, FealOH)^, insoluble in KOH, soluble in HCl.

(NH4)2S precipitates black FeS soluble in HCl, insoluble in KOH.

These tests with the above reactions, produced by K4Fe(CN)g and
KCNS are sufficient to identify ferric compounds in water.

Quantitative Estimation. — If the quantity of iron is small, it may
be estimated colorimetrically like lead and copper, in which case
prepare a standard solution of iron alum, Fe(NH4)(S04)2.i2H20,
by dissolving o-86i gramme in a litre of distilled water. This solu-
tion contains o-oooi gramme Fe per c.c. If necessary, evaporate
half a litre of the water to 100 c.c. Place this in a Nessler glass^
and add i c.c. of dilute K4Fe(CN)g solution. Match the colour of
the liquid in this glass by adding to 100 c.c. distilled water in a
similar Nessler the same amount of K4Fe(CN)g and the requisite
quantity of the standard iron solution. It is well to add a drop or
two of nitric acid free from iron to each of the Nesslers.

If the quantity of iron is too great for colorimetric estimation,
acidify a litre of the water with HCl, and evaporate to dryness.
Complete the drying in the air-bath at 150° C. Moisten with HCl,
add water, and heat. Filter off any insoluble silica (which may be
washed, ignited, and weighed). To the filtrate add a few drops
pure HNO3, and boil. Then add a little (NH4)C1 solution and a
shght excess of NH4OH, and allow the precipitate of ferric hydroxide
to settle. Filter this off; ignite and weigh as Fe^Og.

[Should it be necessary to estimate Ca, which is now contained
in the filtrate, add excess of ammonium oxalate, allow to settle,
filter off, and ignite the calcium oxalate. Weigh as CaO.]

Not more than o-i part Fe per 100,000 should be present in a
domestic water. A distinct chalybeate taste is produced by 0*3
part per 100,000.


Chromium. — Evaporate a litre of the water to be tested to
dryness, and fuse the ash with solid potassium nitrate and sodium
carbonate to produce yellow KoCr04, which, in neutral solution,
produces a red precipitate with AgNOg (soluble in ammonia and
dilute nitric acid), and in. solution in acetic acid gives a yellow
precipitate with lead acetate insoluble in dilute acetic acid. A few
c.c. of a largely concentrated sample may be dropped on a thin layer
of ether which has been floated on a dilute solution of H2O2 acidi-
fied with H2SO4. Upon slight agitation the blue colour which
forms in the lower solution passes to the ether.

In chromates (yellow or red in colour) Cr exists in combination
with oxygen, acting as an acid radicle. Cr also forms a set of salts
in which it acts as a metallic radicle. These are green or violet in
colour, but pass through oxidation into chromates.

Conversely, chromates pass by reduction into green chromic
compounds. Acidify a chromate with HCl, add Zn, and warm;
the yellow chromate passes into a green chromic salt.

(NHJOH and KOH in small quantity produce a pale bluish-
green or purple precipitate of Cr2(0H),j, more or less soluble in excess
of the precipitant.

Quantitative Estimation. — A chromate is first transformed by a
reducing agent into a chromic salt. A solution of the chromic salt
is then precipitated by NH4OH in presence of NH4CI, and the
resulting hydrate converted by ignition into CraOg, and weighed.
From the weight of CroOg the amount of Cr is calculated.

No chromium should be present in a drinking water.

Zinc. — Concentrate the water.

(NH4)2S produces a white, flocculent, gelatinous precipitate,
which often appears yellow owing to excess of yellow ammonium
poh'sulphide, (NH4)2Sa. This reaction is characteristic, as zinc
sulphide is the only white sulphide capable of being precipitated.
Zn is only partly precipitated from neutral solution by HjS, but by
adding sufficient NaOH, NH4OH, or sodium acetate, the whole of
the metal may be precipitated by this reagent.

Solution of NH4OH gives a white precipitate of Zn, (0H)2, readily
soluble in excess of ammonia.

K4Fe(CN)g produces a white gelatinous precipitate of zinc ferro-


Quantitative Estimation of Zn. — Prepare a standard solution of
ZnS04,7H20. In 287 parts of this salt there are 65 of Zn, or in
4-4 parts I of Zn. Dissolve 4-4 grammes of the crystals in a htre
of water (each c.c. = o-ooi gramme Zn). Use this standard solution
volumetrically, as in the case of Fe, precipitating the Zn with
K4Fe(CN)p. Avoid much excess of the ferrocyanide. This may be
effected by placing a drop of the mixture on a white tile in contact
with a drop of a saturated solution of uranium acetate, when a
brown colour appears immediately free ferrocyanide is present.

Gravimetric Estimation. — Heat a measured quantity of the con-
centrated water to boiling, and add slight excess of a solution of
NagCOg. Boil again, and allow the precipitate to settle. Wash
several times by decantation with boiling water; transfer the pre-
cipitate to a filter, and finish the washing thereon. When finished
the wash-water shows no alkalinity to litmus and gives no precipi-
tate with BaClg.

Dry the precipitate, and carefully transfer it to a porcelain cruci-
ble. Heat to redness. Wet the filter-paper with strong ammonium
nitrate solution, and dry it; incinerate it in the flame in a coil of
platinum wire, and let the ashes fall into the crucible. The flame
should not enter the interior of the crucible during ignition, lest
reduction of the ZnO take place. Cool and weigh. Calculate Zn
from ZnO.

Carbonate of Zn, ZnCOg, is found in certain mineral waters in
quantities varying from q-ooi to 0-005 parts per 100,000. As
much as 10 parts per 100,000 ZnS04 have been detected in such
waters. Zinc may be introduced by galvanized iron tanks or pipes.
It should not be found in a drinking water.



As vegetable organic matter has little significance from the sanitary
point of view, attention is almost entirely directed to animal matter
in the form of sewage. It is not proved that animal organic matter
per se in the quantities found even in dilute sewage is hurtful to
health; its importance lies rather in the fact that pathogenic
microbes, especially those of intestinal origin, accompany' it.
Wherever, then, faecal matters in quantity large or small are met
with danger exists.

The complex remains of dead animals and plants are slowly
changed to simple inorganic compounds in the superficial laj^ers of
the soil under the action of manifold ferments, the products of
micro-organisms in association with favourable quantities of heat,
moisture, and oxygen. The sum total of these changes is spoken
of as an oxidation, since the end products are oxides of carbon,
nitrogen, etc. ; but there is no doubt that as in the case of the various
fermentations which take place in the alimentary canal of animals,
known collectively as digestion, reductions frequently alternate
with oxidations. There is some evidence to show that these fer-
ments, metabolic products of aerobic and anaerobic bacteria, act
along certain lines which are intimately correlated. The specific
action of one enzyme furnishes the necessary conditions for the
opposed functions of a succeeding enzyme.

Whilst an accurate qualitative or quantitative estimation of
organic matter in a potable water is impossible, still there are
certain chemical tests of value in directing us towards the source
of the organic matter, which source maj' ultimately be discovered
by other means.

A rough differentiation of animal from vegetable matter may be



effected by a consideration of the ratio of ' organic carbon ' to
' organic nitrogen,' which ratio forms the basis of Frankland's
well-known method of estimating organic matter. The process is
only suitable for experienced chemists and laboratories equipped
with apparatus for gas analysis. But in skilled hands it is simple
and direct. A measured volume of water is carefully evaporated
to dryness; the residue is introduced into a hard glass tube along
with some oxide of copper, and the tube is heated in a furnace
until combustion of the organic matter is complete. The gaseous
products of combustion — carbon dioxide, nitric oxide, and nitrogen
— are severally collected and weighed, as ' organic carbon ' and
' organic nitrogen.' If in surface waters the proportion of organic
carbon to organic nitrogen be as low as 3:1' the organic matter
may be considered as of animal origin, while if it be as high as
8 : I it is chiefly vegetable. In certain fresh peaty waters the ratio
of C : N has been found as high as 12 : i. In fresh sewage the
proportion of C : N may be 2 : i. Frankland held that the smaller
the proportion of organic carbon and organic nitrogen in a water,
and of these constituents the larger the proportion of C : N, other
things being equal, the better is the quality of the water.

The fermentation of dead organic matter, known as ' putrefaction,'
is effected by many types of micro-organisms.

Dead proteins are hydrolysed to proteoses; these to peptones;
peptones to amino-acids; finally amino-acids are split, evolving

If we follow this ammonia as it escapes, say, from, a dung-heap
in solution into the soil, we shall find that in the presence of the
' nitrous ' organisms nitrous acid is formed, which in contact with
the bases of the soil rapidly becomes nitrites. Later, through the
activities of the ' nitric ' group of micro-organisms, nitric acid is
generated, which speedily becomes nitrates. These various stages
in the oxidation or purification of nitrogenous matters stand out as
chemical landmarks, and present considerable information to the
water analyst.

As carbohydrates and fats are much less complex bodies contain-
ing C, H, and only, their decomposition and oxidation are much
more simple: carbon is burnt to COg, and H to HoO.

These changes in nitrogenous matter ma}^ be studied directly.



If, for example, A be a source of organic pollution, say a manure-
heap, on the surface of the ground, and B, C, D, and E wells at
increasing distances from it, analysis will show that the water in
B contains abundance of NH., ; nitrification has not yet taken place.
At C the oxidation processes have advanced to the stage of nitrous
acid; this water will contain less ammonia and some nitrites. The
water from D has travelled farther, encountering more nitrifying
organisms, with the result that ammonia has disappeared, and nitrites
and nitrates are found. At E purification is complete — the whole
of the N is oxidized to nitric acid; hence this water contains no
NHg, no nitrites, but onh' nitrates.

The opportunities for purification offered between A and B are not
sufficient to carry the oxidation changes beyond the stage of NH.,;
whereas the journey from A to E is of such length that the entire



*"" Nitrate

Nitrite Nitrate

Fig. 3.

changes have been completed. At intermediate points are observed
intermediate stages in the purification.

From the consideration of a single instance of this kind, no con-
clusions as to the distance a well must be removed from a source
of contamination in order to be safe can be drawn, since the factors
in the problem of safety are numerous and variable. The distance
between A and E, if the water in E is to be completely purified,
would require to be much greater if the slope from A to E be con-
siderable, or E would need to be much deeper. On the other
hand, if the slope of the ground water descended from E to A, it
is possible that the water of B may be free from all organic matter.
The porosity of the soil, conditions of heat and moisture necessary
to vigorous growth of purifying organisms, direction of slope of
ground water, geological features of subsoil and underlying strata,
rainfall, and a number of other factors, all influence this question
of safe distance of well waters from foci of contamination. Each
case must be worked out on its ow-n merits; and here the chemical


examination renders useful service. A sample of water from E
may be pure to-day — that is, contain no organic matter as such,
no NH3, no nitrites, but only nitrates ; to-morrow, owing to increased
rainfall, whereby more organic matter than usual is washed into
the soil, or to some other condition by which the powers of the soil
for purification are lessened, this same water may contain, besides
nitrates, nitrites, NH3, and even undecomposed organic matter.
It should ever be borne in mind that the machinery by which
organic matter is purified in the soil is liable at any point to break
down, and in too many instances although just sufficient for the
work is near breaking-point. The question, therefore, should be,
not how near to a focus of contamination may it be safe to procure
water, but rather how far from the focus is it possible to acquire
it. Chemical analysis, if frequently and regularly performed, will,
in most cases, discover such breakdown in the purification
machinery, although a single analysis, unaccompanied by further
information as to source and surroundings, may be quite useless.
It is the comparative information regarding a water acquired by
systematic and repeated analyses that is of value.

The student should note that the nearer the nitrogenous organic
matter of domestic sewage stands to the stage of raw proteins
the worse, as it is in this stage that pathogenic bacteria are found
in their most toxic and vigorous condition; and that, conversely,
the farther from this stage such matter stands the less dangerous
it is. When organic matter reaches the stage of nitrates no patho-
genic germs will live in it.

From the standpoint of infection, fresh faecal matter and urine
are the most dangerous of all forms of organic matter.

It is not possible in water analysis to separate and estimate raw
proteins, proteoses, peptones, and amino-acids. The next stage,
that of NH3, lends itself to ready estimation.

When this ' free and saline ' ammonia, as it is called, is removed,
the remaining organic matter, which consists of the nitrogenous
complexes constituting the antecedent stages, can be rapidly
oxidized by the aid of a powerful oxidizer and heat (^\'anklyn's
process) into ammonia, and estimated as ' albuminoid ' ammonia.
This figure, inasmuch as it measures those portions of the nitrog-
enous organic matter likely to contain pathogenic micro-organisms.


is obvioush' the most important determination connected \vith this
portion of tlie subject.

Estimation of 'Free and Saline' NRj. — Prepare a standard
solution of (XHjjCl, i c.c. of which = o-oi milhgramme NH3.

53-5 grammes NH4CI contain 17 grammes NH3.
3-14 ,, ,, ,, I gramme NH3.

Dissolve 3-14 dr}' anhydrous NHjCl in i Htre ammonia-free
distilled water. One c.c. of this solution =1 milligramme NH3.
This is too strong. Dihite 10 c.c. of it to a litre; i c.c. now=o-oi
milligramme XH3.

The process depends on the fact that when the water is distilled
with a little sodium carbonate all the ammonia in the water, free or
combined, passes over in the first portions of the distillate, and may
be estimated by Nessler's solution.

Prepare Nessler's solution. Dissolve 62'5 grammes KI in about
250 c.c. distilled water. Set aside a few c.c. of this solution. Now
add to the larger portion saturated mercuric chloride solution till
precipitated mercuric iodide cea,ses to dissolve on stirring. Add
the reserved KI so as to redissolve the precipitate, and again add
cautiously sufficient mercuric chloride solution to produce a shght
permanent precipitate.

Dissolve 150 grammes KOH in about 300 c.c. water; cool; add
gradually to the above solution, and make up with HoO to a litre.
A brown precipitate settles out on standing, and the supernatant
fluid is clear and of a pale greenish-yellow colour. It is ready for
use as soon as it is perfect!}' clear. It should be decanted without
stirring up the sediment. Keep in bottles closed with well-fitting
rubber stoppers. This solution is rendered sensitive from time to
time by the addition of a little more HgCU solution ; its sensitiveness
depends on its being saturated with HgCU.

Sodium Carbonate. — Heat anhydrous NaaCOg to redness, taking
care not to fuse it ; transfer to a mortar, and grind to a fine powder.
Store in a clean, dry, wide-mouthed, stoppered bottle.

Ammonia-free water is prepared bj' distilling ordinary water in
the presence of NaaCOg or H.,S04, and rejecting the first portions of
the distillate until there is no trace of colour produced on Nesslerising
50 c.c. of it.


A preliminary test may be made in order to ascertain what
quantity of the water-sample should be distilled in order to make
an exact determination of the ammonia. Place two Nessler glasses
on a white tile; add 50 c.c. of the sample to one, and 50 c.c. am-
monia-free distilled water to the other. To the ammonia-free
water add 0-5 c.c. of the dilute standard NH4CI solution. To both
Nesslers now add 2 c.c. Nessler's reagent, and stir. If on standing
five minutes the intensity of colour in both cylinders is the same,
500 c.c. of the water may be used for distillation. If the intensity
of the colour of the sample is much greater, dilution is necessary
prior to distillation, otherwise the quantity of ammonia in the first
50 c.c. distillate will be too large to match.

Arrange a distilhng-flask, condenser, and Bunsen burner. Pour
into the flask 500 c.c. of the water (or water sufficiently diluted);
add some prepared sodium carbonate, and if the water is acid a
little more than usual (the least acidity fixes NH3). Receive in
Nessler glasses 150 c.c. distillate in three lots of 50 c.c. each. The
boiling should be briskly effected; it is generally useful to place a
piece of pumice in the flask to prevent bumping. As each Nessler
glass is filled it should be Nesslerised or covered until Nesslerisation
is accomplished.

Nesslerisation is one of a number of colorimetric methods of
volumetric analysis in which the amount of a substance is esti-
mated by adding to it a second body capable of forming a char-
acteristic colour with it. The same conditions are accurately
fulfilled in a similar vessel, using distilled water and such quantity
of the substance sought, in standard solution, as will match the
colour of the first when the same quantity of the second body is
added. In order that shght differences in tint may be appreciated
and matched, it is necessary to work with dilute solutions of the
body to be estimated and the standard reagent ; hence the necessity
at times of diluting the water under examination.

Having collected the three 50 c.c.'s of distillate, Nesslerise each
separately. Stand the Nessler glass on a white tile in a good north
light, and by its side place a second Nessler glass of similar shape
containing distilled ammonia-free water, and that quantity of
standard solution of (NHJCl deemed necessary to match the first.
Into each deliver 2 c.c. of Nessler's reagent, and carefully mix. In


a few minutes the yellow colour will have fully developed, and its
depth can be gauged by looking down through the column. Should
there be some discrepancy in the tints, rapidly add to another
Nessler glass containing distilled ammonia-free water a little more
or a little less of the standard solution, as the case may be, until
an exact match is produced. In all such colorimetric work every
condition should be exactly similar in the two cases — length of
time reagents are in contact, order in which reagents are added,
shape and size of containing vessels, etc. The standard solution
of XH4CI must be added to the second Nessler glass before the
Nessler's reagent, as this occurred in the Nessler glass containing
the distillate. If the standard solution be added after the Nessler
reagent an opacity is likely to form which prevents to some degree
an exact match being made. Several trials may be necessary before
an accurate result is reached.

The second and third 50 c.c. of the distillate are treated in the
same way, and the sum of the results in terms of c.c. of the standard
solution noted.

\\'anklyn found that the whole of the free and sahne NH3 was
contained in 150 c.c. distillate, and that the first 50 c.c. contained
three-fourths of the total.

Nesslerise the second distillate first, and note whether more then
1-5 c.c. of the standard NH4CI solution is required to match it. if
so, the first distillate must be diluted before Nesslerisation, other-
wise the colour will be too intense to be accurately matched.


First Nessler glass matched by 3-00 c.c. XH4CI (ic.c. = o-oi milligramme XH3)
Second ,, ,, ,, ,, 0-75

Third ,, ,, ,, ,, 0-25

Total XH3=4'Oo ,, ,, ,, ,,

But each c.c. standard NH4Cl = o-oi milligramme NH3;
.■ . 4 c.c. = 0*04
And in 500 c.c. of the water under examination there is 0*04 milligramme NH3.
In 100 c.c. there will be o-ooS milligramme XH3, or, since 100 c.c. water =
100,000 milligrammes, this water contains free and saline XH3 to the extent of
O'OoS part per 100,000.

Estimation of 'Albuminoid' NH3.— \Miilst the Nesslerisation
of the free and saline NH3 is going on, 50 c.c. of alkaline potassium
permanganate (composed of 200 grammes KOH, 8 grammes per-


manganate, a litre of water) should be boiled, so as to expel any
ammonia that it may contain, and to heat the liquid in order to

Online LibraryDavid SommervillePractical sanitary science : a handbook for the public health laboratory → online text (page 4 of 27)