David Sommerville.

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

. (page 2 of 27)
Online LibraryDavid SommervillePractical sanitary science : a handbook for the public health laboratory → online text (page 2 of 27)
Font size
QR-code for this ebook

of the analysis, should be transcribed into a book and preserved for
future reference.

Estimation of Odour. — Place 250 c.c. in a stoppered flask,
and heat to 37° C. in an air-bath. Remove the stopper and
smell. It is generally sufficient to shake the sample well in the cold,
rapidly remove the stopper, and smell. The variety of odours is
inhnite. Many odours are produced by organisms, either as products
of their life-history or of their death and putrefaction. Beggiatoa,
Chara, and certain species of Crenothrix produce an offensive odour
of H.,S. It is believed that Beggiatoa during its life-cycle reduces
sulphates, and produces under favourable circumstances large
quantities of H.,S. Crenothrix, moreover, often produces abundance
of colour, varying from brown to red. Tabellaria, Meridion, and
certain diatoms, as also the protozoon Cryptomonas, furnish a dis-
tinctl}' aromatic odour. A lishy odour is produced by Volvox and
the protozoa O^cnodinium, Bursaria, and Uroglena. A grassy
odour accompanies Rivularia, Anabsena, and Caelosphserium.

Taste. — Pure rain water well aerated has a fairly distinctive
taste, more easily appreciated than described. So also have peaty
waters, sea water, and chalybeate waters. The taste of a particular
sample may be, however, everything to be desired, whilst the water
is the foulest of the foul. Taste, however, is of little service to the
analyst, and not always to be recommended. Iron is about the


only ingredient that can in this way be detected in very small quan-
tities, being recognisable to the amount of 0-5 part per 100,000.

Potable waters have been classified as —

Wholesome.— {1) Spring water; (2) deep- well water; {3) upland
surface water.

Suspicious.— {^) Stored rain water; (5) surface water from culti-
vated land.

Dangerous.— [6) River water polluted with sewage ; (7) shallow-
well water.



It will be well for the student from the lirst to ht up his own appa-
ratus, and make his own standard solutions. He must learn to
use properly the chemical balance, and a special demonstration is
devoted to the mechanism, methods of adjusting and using this
all-important instrument. Before commencing to weigh, he should
see that the balance is accurately levelled, and that the index moves
without effort over the whole field of the graduated scale, and comes
to rest at zero. All weights, basins, etc., should be transferred to
and from the scale-pans only when these are supported. It is
customary to use three rows of weights, grammes (brass), deci-
grammes and centigrammes (platinum), and milligrammes (plati-
num). A rider of platinum applied to the beam also reads milli-
grammes. The right-hand pan should be used only for weights,
and these should be placed methodically in three rows in front of the
operator. By this means the total reading is most easily obtained
and checked.

Immediately on finishing a weighing all weights should be trans-
ferred to the box, with the forceps used for the purpose, and the box
and balance carefully closed.

The standard solutions in use in water analysis are of two types :

1. Normal, decinormal, centinormal, etc.

2. Standards of such strength that a litre contains the equivalent
of a gramme, or submultiple of a gramme, of the substance to be

A standard solution is said to be normal when one litre contains
the equivalent weight in grammes of an element, acid, alkali, or

The molecular weight of HCl is 36-35; therefore 36-35 grammes
HCl per litre = normal HCl, written N.HCl.


In like manner N.NaOH= 40 grammes per litre.
Since the term ' equivalent ' signifies the weight in grammes of the
substance under consideration, which is chemically equivalent to

1 gramme of H, normal H2S04= =49 grammes per litre.

A decinormal solution {—) is one-tenth the strength of a normal
— thus 3~j- NaOH=4 grammes per litre — and a seminormal (^) and
centinormal (y^jj) are respectively one-half and one-hundredth the
strength of the normal — viz., 20 grammes and 0-4 gramme per litre

In tribasic acids one-third of the molecular weight in grammes
per litre constitutes a normal solution, and so on for acids of higher

The terms ' normal,' ' decinormal,' etc., are used sometimes with a
different meaning. Permanganate of potassium, as we shall see
presently, in acid solution is reduced by many substances, accord-
ing to the equation KgMn.Pg = KgO + 2MnO -t- O5, in which

2 gramme molecules of KMn04 correspond to 5 gramme molecules

of oxygen or to 10 gramme molecules of hydrogen. Accordingly,

in order to put permanganate of potassium on a hydrogen basis, a

1 1 X- ■ J . X ■ 316-3 /2KMn04\ r

normal solutionis made to contain - — - = 31' 03 grammes

10 V 10 / ^

per litre. In the same way K2Cr207. which in acid solution parts
with O3, requires for a normal solution -^ — ( ^a^ ^ ) = 49'^

grammes per litre.

The second type of standard solution used is constructed so that
a minimum amount of calculation suffices in estimating results.
Since it is customary to represent the various items of the analysis
as parts by weight per 100,000 of the water, and since i c.c. of water
weighs I gramme (1,000 milligrammes), 100 c.c. of water will weigh
100,000 milligrammes.

It is therefore convenient, when possible, to work on 100 c.c. of
the water sample throughout the various estimations, and to use a
standard solution that will give readings directty in the above

Suppose we wish to estimate the quantity of CI in a water, we
use a solution of AgNOg of such strength that i c.c. is equivalent


to I milligramme CI. To make this solution we refer to the molec-
ular weight of AgNOg, and the atomic weight of CI. AgNOa-t-
XaCl-AgCl + NaNOg.

170 grammes AgNOg precipitate 35-35 grammes CI.
.-. dividing by 35-35, we find that 4-8 grammes AgNO., precipitate
I gramme CI.

If, then, we dissolve 4-8 grammes AgXO.j in i litre of water we
obtain a solution i c.c. of which precipitates i milligramme of CI,
and working with 100 c.c. of water, the number of c.c. of the
silver nitrate solution used indicates the number of milligrammes
of CI in 100,000 milhgrammes of the water, which is parts per

Again, in estimating XH3 a standard solution of XH4CI is pre-
pared and used in the same way.

One molecule of XH4CI contains i molecule of XH^.
53*35 grammes ,, contain 17 grammes
and 3-14 ,, ,, ,, I gramme

Therefore a litre containing 3-14 grammes XH4CI will contain
I gramme NH3, and consequently i c.t. contains i milligramme.
It is found convenient to dilute this 100 times, so that i c.c.= o-oi
milligramme XH3.

Standard solutions should be stored in bottles in such manner
that both internal and external evaporation are impossible. In
the first case, where the bottle is not quite full, pure water will
evaporate and condense on the upper portions of the vessel; in the
second, evaporation will take place into the atmosphere. The loss
of water will naturally depend on the substance dissolved, the tem-
perature, the age of the solution, and the frequency with which it
is used. A rough estimate may often be made of the probable
amount of change in strength by noting the date of preparation,
which should always be found on the label. Some standards undergo
chemical change by the action of light, and should therefore be kept
in the dark.

In reading a burette, arrange it so that the lower convex line of
the meniscus is in the same horizontal plane with the eye; the


division of the scale cut by the lowest point of this convex line is
the reading.

In measuring small quantities of liquids much time may be saved
by using a few plain 10 c.c. pipettes graduated to tenths of a c.c, and
for quantities under a c.c. a i c.c. pipette graduated to hundredths.
These can be easily and rapidly cleaned, and as easily and rapidly
manipulated, and may often take the place of burettes. In
weighing platinum and porcelain basins, crucibles, etc., it is very
necessary to see that they are quite dry. To insure this, especially
after heating, they should be placed for ten minutes in a desiccator
immediately before going to the balance. It is also necessary to
be certain that all such vessels are thoroughly clean. Accurate
notes of all operations, measurements, weights, etc., should be made
in the bench notebook, and considered as much a part of the work
as the operations themselves. Without this notebook it is impos-
sible to get on with analytical chemistry. Where possible it is well
to write down the chemical equations representing decompositions.
When in doubt in this matter refer to a work on chemistry. All
colour matches are best made in glass cylinders standing on a white
ground, as the operator faces a north light.

The Reaction of Water. — This is an important item, and
should form the first step in the routine chemical examination. In
addition to the use of red and blue litmus-papers, it is often well to
use a more delicate indicator, such as phenolphthalein, and to esti-
mate the amount of acidity (when acid) in 100 c.c. by titrating with
-^0 NaOH, or of alkalinity (when alkaline) with ^ H2SO4. An acid
water dissolves lead, iron, and zinc; it also fixes ammonia, and so
prevents its being distilled off. Some hold that neutral waters and
those possessing very slight temporary hardness are capable of
dissolving lead. It should be remembered, however, that sodium
carbonate when present prevents this action. Houston has corre-
lated the acidity and plumbo-solvency of a large number of moor-
land waters.

He causes the sample to percolate upwards through a column
of specially prepared lead shot at a uniform rate. He then
collects successive 50 c.c.'s, and estimates the amount of lead in

The following figures are taken from a report to the L.G.B.:





Number of c.c.


ms. of I'b in

N Na.iCOj required to neutralise

100 c.c.

of the Water after

" loo c.c. of the Water.


through Lead Shot.





















Some waters not acid, and failing to dissolve lead, exert an
' erosive ' action, forming an insoluble film of oxyhydrate upon
the lead, which after a time may become detached, and produce a
degree of opacity.

Chlorides in Water. — Free CI rarel}' occurs in water-supplies.
Certain manufacturing effluents ma3' on occasion contain small
quantities of free CI, but the quantit}^ is so small and the occurrence
so rare that this form of CI may be practically ignored. The great
bulk of CI in drinking water is found as NaCl. All soils and sub-
soils contain this salt in large amounts. The water-bearing strata
are rich in chlorides, especially NaCl, and consequently rain water
(which itself may contain as much as 0-5 part per 100,000 NaCl),
as it percolates from the surface to the impermeable stratum on
which it rests, dissolves these in considerable quantities. CaClg
and MgClg are found in certain strata— chalk and limestone — in
much smaller quantities, but MgCU abounds in sea water, and in
large quantity is distinctive of it. Wells, reservoirs, etc., to which
sea water can obtain access will yield waters rich in ^MgCl,. Sources
of water subject to much evaporation, especiall}' if situated near
the sea, exhibit large quantities of chlorides. The total CI in sea
water approaches 2,000 parts per 100,000, and if this figure be kept
in memory it will explain the large estimations often found some
considerable distance from the littoral. During the passage of water
through the soil, subsoil, and strata, CI is not likely to be diminished
as are the organic matter and bacteria.

When we have accounted for all the CI contributed by rain water,


sea water, soil, subsoil, and strata, and trade effluents from chemical
works, paper factories, etc., there may remain a surplus furnished
by organic pollution of animal origin. This surplus is of some
import to the analyst, as indicating sewage; but before it is returned
as such all the possible sources of origin just mentioned must be
rigidly excluded. Vegetable organic matter does not yield this
surplus CI. Attempts have been made in U.S.A. to estimate and
permanently record the CI due to the natural causes named, so that
sewage pollution may be readily detected. Maps have been con-
structed and points furnishing equal quantities of CI joined by lines
named ' isochlors.' In districts remote from the sea, and centres
of population and land cultivation, such maps may be more or less
reliable, but in this country they would be useless. MHiilst it is
true that animal pollution contains much CI (urine about i per
cent, chlorides), and that soils, strata, etc., in certain districts yield
fairly constant quantities, still there are variations in many localities
in these natural sources, and it is only where large quantities of sew-
age have gained access to waters that we can rely on the surplus
CI as evidence of this accession. In the case of small amounts of
sewage this surplus CI figure is of little if any value. But in a water
analysis the most important information lies very often not so much
in the exact amount of a particular constituent as in the fact that its
presence points to past pollution, and consequently to the possibility
and even probability of a recurrence of such pollution. In this
light CI and nitrates play an important role. These afford unmis-
takable evidence of previous contamination; they are the distinct
and unchangeable indications of previous pollution, but as to
whether recent or remote they indicate nothing. Hence the neces-
sity for further and different forms of examination. As to the
amounts of chlorides that should condemn waters, it is difficult to
speak, since there is such infinite variety in the quantities contained
in different soils and strata. MgClg and CaClg render waters hard,
so that more than 4 or 5 parts of either or both of these per 100,000
will cause a large destruction of soap, and these figures will in most
cases form the limit for domestic waters. NaCl may go up to
perhaps 50 parts per 100,000; above this it imparts a taste, and the
water consequently will not be fit for drinking.


Estimation of CI.

Apparatus and Reagents Required.

A white porcelain basin capable of holding 250 c.c.

A glass stirring-rod.

A burette charged with standard solution of AgNOa, of which
1 c.c. is equivalent to i milligramnie CI (4"8 grammes AgNO., to a
litre of water).

A 5 per cent, solution of KoCrOj.

Place 100 c.c. of the water in the porcelain dish.

Add I c.c. of the K2Cr04 solution, and stir.

J\un in from the burette drop bj^ drop the silver nitrate solution
until the pale ^-ellow colour remains permanently orange.

Take the reading.

The rationale of the process is as follows:

AgNOg, when added to a solution of chlorides, forms AgCl, a
white curdy precipitate insoluble in HNO3, soluble in NH^HO.
Without a special indicator it would be impossible to determine
when the whole of this white precipitate had been formed — when the
whole of the CI had been deposited.

K2Cr04 is also acted on by AgNOg, and Ag2Cr04 formed, which is
red. But so long as any chloride remains ununited with Ag, the
silver chromate is decomposed and AgCl formed; hence the dis-
appearance of the red colour on stirring. Immediately the whole
of the CI is precipitated as AgCl the red silver chromate remains.

The reactions are represented by the equations —

AgNOg + NaCU AgCl + NaNOg.

2AgX03 + K.,Cr04 = Ag2Cr04 + 2KXO3.

2NaCl + Ag2Cr04=2AgCl + Na2Cr04.

It is obvious that the K2Cr04 should be free from CI. Acidity in
the water will dissolve Ag2Cr04 ; hence if a water is even slightly acid
it must be neutrahzed. Freshly precipitated CaCOg is the best alkali
to use, and it should be used only to the point of neutralization.
If too little K2Cr04 is used the CI reading will be too high, and if
too much be used it is difficult to determine the end; i c.c, accord-
ingly, is found a suitable quantity when the solution is of the above


strength. It will be noticed that as the titration proceeds the red
AggCrO^ disappears more slowly on stirring, until finally it ceases
to disappear. This is explained by the, continuous decrease in the
original chloride. Whilst abundance of this undecomposed chloride
remains in solution, the Ag2Cr04 is rapidly robbed of its Ag and the
red colour discharged; but as the chloride diminishes and the end
approaches, the decomposition of the Ag2Cr04 becomes slower and
slower, until at the end of the reaction it ceases, and the red
Ag2Cr04 permanently remains.

Since the colour-change from pale yellow to red is somewhat
difficult to detect in daylight (it is more easily perceived by gas-
light), a flat glass cell whose plates are | inch apart should be filled
^dth chromate solution of the same tint as that of the contents of
the basin, and interposed between the eye and the basin during
titration, when the appearance of the red silver chromate becomes
strikingly manifest. The effect is to neutralize the yellow and to
cause the appearance of the basin to be the same as if it were filled
with pure water. In working with turmeric, cochineal, etc., cells
should be used filled with corresponding solutions of turmeric,
cochineal, etc.

The number of c.c. of silver nitrate run in represents the number
of parts of CI per 100,000.

100 c.c. water = 100,000 milhgrammes,
I c.c. AgN03= I milhgramme CI;
.■. the number of c.c. AgN03 used= number of parts CI
per 100,000 water.

Some operators subtract -i c.c. from the AgNOg figure as the
quantity required to form the slight permanent orange colour.
Others add a small measured quantity of the water sample from a
burette until the permanent orange tint departs, and reckon half of
this with the AgNOg reading.

It is well always to do tw^o careful estimations, and take the mean.
When once an idea of the quantity of CI present is obtained, two
careful estimations can be performed very rapidly. A control basin
containing 100 c.c. of the same water and i c.c. of K2Cr04 may assist
ill determining the end reaction.

Where small quantities of CI are to be estimated, 250 c.c. or 500 c.c.


of the water may be concentrated by evaporation to lOO c.c.
Alkaline silicates, nitrates, and phosphates slightly affect the CI
estimation, but not to such a degree as to require correction.

Chlorine is sometimes returned in terms of sodium chloride

This figure is found bv multiplying the CI return by - — ~ . WHiere

CaClo, or i\IgClo, or both, enter into the problem, corrections have
to be made in accordance with the respective molecular weights
and the quantities of each present.

In chalk and red sandstone waters 3 parts of CI per 100,000 may
occasion no suspicions of sewage, and 4 or 5 parts may be passed,
unless organic pollution is indicated by other items of the analysis.
Pure surface waters seldom contain more than i part per 100,000,
whilst deep greensand waters may give rise to 15 to 20 parts per
100,000, and still be absolutely pure.

The following are a few examples of the CI figures for different
waters :

per 100,000.

A well in St. Pancras - - - 4-5

Lambeth water-supply - - - - - i -g

., ,. - - - - - 2-0

Southwark water-supply - _ . . - 1.85

A well in Devonshire - - - - - 3-1

Thames water at Waterloo Bridge - - - lo^-z

Deep well near Hindhead _ _ . - ii2'3

Sample of rain water taken from rain gauge in Herts - 0-3


The hardness (soap-precipitating power) of a water exerts little
influence on health, but from an economic point of view is of some

A soap is a chemical salt formed by the union of an inorganic
base with one or more fatty acids.

Sodium and potassium soaps are soluble in water, and when
shaken with it form a dense froth or lather. Calcium and mag-
nesium soaps are insoluble in water, and fail to form a lather. Hence,
if a solution of a soluble soap be added to water containing calcium
or magnesium salts, these last will be completely precipitated in
the form of insoluble calcium or magnesium soaps before a lather
is produced. Accordingly, by using a standard soap solution, an


approximate estimate of the quantity of such soap-precipitating
bodies in a water can be made. The total quantity of such bodies,
as measured by the standard soap solution, constitutes the total
hardness. Other bodies than calcium and magnesium salts are
occasionally present in water, which act in a similar manner on soap.
If much sodium chloride be present, it will precipitate soap from its
solution in an unaltered state.

CaCOg and MgCOg, especially the first, have by far the greatest
share in rendering waters hard. These salts are formed in solution
in the soil as bicarbonates [Ca(HC03)2 and Mg(HC03)2] by COg
dissolved in rain water. On boiling such waters, CO2 escapes, and
insoluble carbonates separate out as a precipitate —

Ca(HC03)2-^CaC03 + CO2 + HgO.

The addition of slaked lime to water containing the bicarbonates
of the alkaline earths results in the precipitation of the lime added
and the bicarbonates thus :

Ca(HC03)2 + Ca(OH)2= 2CaC03 + 2H20. (Clark's process.)

If now the boiled water be filtered, made up to its original volume
with distilled water, and again titrated with standard soap solution,
the permanent hardness is obtained.

The difference between the total and the permanent hardness is
the temporary hardness.

The soap test has been made to measure the quantity of CaCOg
and other salts which produce hardness, but this is not accurate
quantitative analysis. It should be ciearly understood that the
chemical action is multiple and indefinite, and altogether different
from that which usually takes place, when in quantitative analj'sis
we titrate one definite compound against another. All that can
be claimed for the soap process is that it indicates the amount of
soap-destroying bodies present in a given water, but fails to form
a measure for any in particular.

The following compounds produce hardness:

CaCOg, MgCOg, CO2 in solution, CaS04, MgS04, FeoOg, and other
Peroxides, zinc salts, Si02, A\^ (OH)e, chlorides, nitrates, phosphates,
and free mineral and organic acids.

The temporary hardness, which is got rid of by boiling, is fcr


the most part produced by CaCO^ and MgCOg, held in sokition by
COo. After these come small quantities of CaS04 and MgS04,
which are also thrown out immediately C0.> is driven off, but the
great bulk of these sulphates remains in solution. Lasth', in a
few cases minute quantities of oxides of Fe, silica, and alumina
are deposited. Phosphate of Ca, if present in appreciable quantity,
may, under certain conditions, be deposited in very small amounts.
On cooling some of the precipitated MgC03, and to a less degree
CaCOg, CaS04, and Ca3(P04)., will redissolve and go to form per-
manent hardness.

MgCOg destro^'S nearly 50 per cent, more soap than CaCOg, but
is found in potable waters in very much less quantity.

Estimation of Hardness. — Prepare a standard solution of
calcium chloride in the following manner: Weigh accuratel}^ 0-2
gramme pure calcite (CaCOg), and dissolve it in dilute HCl, taking
care to keep the vessel covered so as to avoid loss by spirting. Evap-
orate this solution to dr3mess on the water-bath. Add water, and
again evaporate to dryness, and repeat these processes in order to
remove all free hydrochloric acid. Now dissolve the residue of neutral
CaClo in water and make up to a litre. One c.c. = the equivalent
of 0-2 milligramme CaCOg. In other words, this solution possesses
hardness = 20 parts per 100,000.

Prepare a standard soap solution by dissolving about 13 grammes
of Castile soap in a litre of equal parts methylated spirit and water.
Stand in a cool place for some hours, and filter.

The titration and dilution of this soap solution is carried out as
follows :

Make up 50 c.c. of the calcium chloride solution to 100 c.c. with
distilled water (10 parts hardness per 100,000), and place in a

Online LibraryDavid SommervillePractical sanitary science : a handbook for the public health laboratory → online text (page 2 of 27)