HEFORT 
ON THE 
Purification of Drinking, Water by Alum*' ' 
By Profs. Peter T. Austen, PhD.. F.C.S., 
and Francis A. Wilber, ftT.K 
- ' 
The many discoveries that have 
been made during the last years in re- 
gard to the transmission of diseases by 
drinking-waters have caused attention 
to be directed to the methods of its ex- 
amination and the processes for purify- 
ing it. Chemical analysis can establish 
the presence of albuminoid matter in 
water, and by its means we are able to 
state if the water under examination 
can'become a suitable nidus, or me- 
dium, for the development of 
disease germs. If the germs are actually 
there, or if the water contains a virus, 
or ptomaine,f biological examination 
alone can determine. 
While physicians and scientific men 
are experimenting on the methods of 
water examination, and are endeavor- 
ing to understand fully the meaning of 
the results obtained, the public are 
chiefly interested to have some method 
by which they can purify their drink- 
ing water in a simple, cheap, efficacious, 
and expeditious manner. 
Running over the substances which 
have been suggested and tried for the 
purification of water, there is none that 
seems to offer the advantages of alum. 
Particular attention was directed to its 
use by Jeunet in 1865, in an article 
published in the Moniteur Scientifique 
(page 1007). He found that 04 gramme 
of alum to a litre of 1 water (23'3 
grains to one gallon) rendered it drink- 
able, even when it was quite full of 
foreign matter. The time taken for 
this clarification was from seven to 
seventeen minutes. 
Alum is a double sulphate of potash 
and aluminium, and in this case breaks 
into potassium sulphate, which remains 
in solution, and a basic aluminic sul- 
phate. This basic sulphate of alumin- 
ium, the composition of which is unde- 
termined, precipitates as a more or less 
gelatinous and flocculent mass, and 
carries down with it the foreign matters 
and humus bodies. The sulphuric acid 
set free in the formation of the basic 
aluminic sulphate attacks the earthy 
and alkaline carbonates, whichfare al- 
ways present, and forms with them 
sulphates, setting carbonic acid free. 
Aluminic sulphate acts like alum. Alu- 
minic acetate and ferric acetate do 
not give such good results. Peligot, 
however, found that ferric chloride, 
(sesquichloride of iron) acted well, and ' 
Scherer recommends the use of a neu- 
tral sulphate of iron. 
In the last years an extensive use of ! 
alum has been made in the many pro- 
cesses of purifying water, sewage, etc. 
It is not improbable that aside from its 
I effect in precipitating matter' mechan- 
ically by envelopment withthe precipi- 
1 tating' basic aluminic sulphate, the 
alum exerts a distinct coagulative ac- 
tion on the albuminous substances in 
the water, rendering them insoluble, 
. and thus causing their precipitation: 
, perhaps the same or similar effect that 
alum produces in the tawing of leather. 
One of the most prominent applications 
of alum at present is in the Hyatt filter. 
By the addition of a minute amount of 
alum, water is rendered capable of a 
most perfect mechanical filtration. 
The fact that alum is cheap, and can 
be obtained in quite a pure state at 
any drug store, places it within the 
reach of every one. Its sharp taste pre- 
cludes the possibility of its being swal- 
lowed by mistake. But even should it 
be swallowed by mistake, no great 
harm would be likely to ensue, unless a 
large amount were taken. In medical 
treatment as high as 30 grains are given 
in a single dose, and this may be re- 
peated four times per day. If it can 
be proved that alum not only clarifies 
a water, but also removes from it dis 
ease germs and ptomaines, its use will 
prove of incalculable value to the hu- 
man race, for facts begin to indicate that 
a vast number of diseases are communi- 
cated through drinking water. 
The investigation of the effects of 
alum on drinking water falls under the 
several heads- 
I. Clarification of the water by set- 
tling. 
II. Clarification of the water by fil- 
] tration. 
III. Use of water clarified by alum in 
manufacturing. 
IV. Removal of disease germs. 
V. Removal of ptomaines. 
VI. Removal of organic matter. 
The investigation must needs be both 
chemical and biological. Only the first 
and part of the second cases have so 
far been examined. 
I. THE EFFECT OF ALUM IN CLARIFY- 
ING WATER BY SETTLING. 
It is evident that to obtain practical 
! results in the clarification of water by 
alum, it must be added in such small 
' amounts as to leave no unnecessary 
excess, and that neither taste nor phy- 
siological action should be imparted to 
; the water. At the time of our experi- 
ments (January, 1885) the New Bruns- 
wick city water was quite turbid from 
' clayey and other matters, so that we 
were able to obtain some very reliable 
results. 
The amount of alum used in the ex- 
periments of Jeunet seems to be unnec- 
essarily high, in case the water is to be 
। drunk. Water was treated with the 
j amount of alum recommended by 
* From the advance sheets of the Annual Re- 
port of the State Geologist of New Jersey for 
1884. 
t Putrefaction alkaloid. 2 
Purification of Drinking Water by Alum. 
Jeunet (23'3 grains to the gallon), but 
no perfect settling was obtained under 
six hours or more ; in some cases not 
under twelve hours. The water thus 
treated had no perceptible taste of 
alum, but it gave a decided reaction 
for alumina when treated with am- 
monia, showing that the water con- 
tained a certain amount of free alum. 
While the amount is evidently too small 
to produce any physiological effect, 
there seems to be no necessity to use 
such an excess. 
To determine the effect of alum as a 
precipitating agent, tall cylinders were 
filled with water and a solution of alum 
was added, the whole well mixed, and 
allowed to stand. It was found that 
in varying lengths of time, depending 
on the amount of alum used, a gela- 
tinous precipitate settled out, and the 
water above it became perfectly clear. 
On adding a relatively large amount of 
alum, and mixing, the coagulation and 
separation of the precipitate is at once 
visible, the water appearing by careful 
examination to be filled with gelat- 
inous particles. The amount of alum 
necessary for the precipitation of a 
water will, of course, depend on the 
amounts of impurity present, but in 
the present case, which may be taken 
as a typical one, we found that 0'02 
gramme of alum to a litre of water (1'2 
grains to a gallon) caused the separa- 
tion and settling of the impurities, so 
that the supernatant water could be 
poured off. This amount of alum was 
shown by numerous experiments to be 
about the practical limit. The com- 
plete settling took place as a rule in 
not less and usually more, than two 
days. It is evident that the amount 
of alum thus added is too slight to be 
perceptible to the taste, and can exert 
no physiological action. We were un- 
able to detect the slightest taste or 
change in the water so treated. 
Still smaller amounts of alum will 
produce a precipitate after longer stand- 
ing. Sixty litres of the city water were 
treated with two grammes of alum (this 
was about 31 grains to 16 gallons) and 
allowed to stand. After forty-eight 
hours the precipitation seemed com- 
plete, and the water was perfectly 
clear, while the bottom of the vessel 
was covered with a brownish, slimy de- 
posit. This substance was collected, 
dried, and analyzed. It gave- 
Carbon 16'50 per cent. 
Hydrogen 2'02 " 
Nitrogen 0'77 " 
It is evident from this analysis that 
a large amount of the organic matter 
has been removed from the water by । 
the alum treatment. 
On incineration, it yielded 59'28 per 
cent, of ash, which contained silica 
and alumina in relatively small 
amounts, oxide of iron in large 
amounts, and a considerable quantity 
of phosphoric acid. 
To determine if there was free alum 
in the water, a sample of the clear 
water, filtered off from the precipitate 
produced by the alum, was made 
slightly alkaline with ammonia and 
warmed for some time. Only the 
merest traces of an alumina reaction; 
could be obtained, and, in fact, in 
some cases, it was doubtful if a reac- 
tion was observable. To prove that 
no more matter could be precipitated 
by the addition of a greater amount of 
alum, samples of the clean filtered 
water were treated with more alum, 
but there was in no case any indica- 
tion of further precipitation on 
standing. 
We consider it, then, established 
that, by the addition of two grains of 
alum to the gallon, or half an ounce to 
one hundred gallons, water can be clari- 
fied by standing, and that neither 
taste nor physiological properties will 
be imparted to it by this treatment. 
By increasing the amount of alum, the 
time required for the separation and 
settling can be diminished, and vice 
versa, by diminishing the amount of 
alum added, a greater time will be re- 
quired for the clarification. 
This method is particularly adapted 
to the clarification of large volumes of 
water, where filtration is not practi- 
cal. The cleared water can be racked 
off to as low a level as possible, after 
which the sediment should be washed 
out and the receptacle cleansed by a 
free use of water. 
II. THE EFFECT OF ALUM IN CLARI- 
FYING WATER BY FILTRATION. 
In order to test the clarification of 
water by filtration after addition of 
alum, the New Brunswick city water 
was again made the subject of'our ex- 
periments. It was found that the sus- 
pended clayey matters were so fine that 
the best varieties of filtering papers were 
unable to remove them. Even when 
several layers of heavy Schleicher and 
Schiill paper were used, a very large 
portion of the suspended matters passed 
through. This, however, is not sur- 
[ prising, since it is well known that the 
mineral matters suspended in water 
are of a remarkable degree of fineness. 
Thus the water of the river Rhine, near 
Bonn, cannot be clarified by simple 
filtration, and takes four months to 
settle. The addition of certain chemi- 
cals aids the filtration of suspended 
matters in some cases, but it does not 
always entirely remove them. Cal- 
cium chloride and other salts are 
recommended as effective agents in 
aiding the removal of suspended mat- 
ters, but in the case of New Brunswick 
water, at least, they have no apparent 
action. The following substances 
were found to have no effect in aiding 
the filtration of the water : sodium 
salts-chloride, carbonate, nitrate, acid 
j carbonate, hydrogen phosphate, acid 
' sulphite, ammonium phosphate, sul- 
phate, biborate, tungstate, acetate; 
potassium salts-hydroxide, chloride, 
bromide, iodide, acetate, phosphate; 
ammonium salts-chloride, sulphate, 
nitrate, acetate ; calcium salts-oxide, 
chloride, sulphate, nitrate. Zinc sul- 
plate and ferrous sulphate (copperas) 
had no action. Acid sulphate of po- 
tassium and of sodium had a slight 
clearing action. Acetate and chloride 
of zinc had an apparent action. Fer- 
ric chloride (perchloride of iron) clear- Purification of Drinking Water by Alum. 
3 
ed perfectly, as also did the nitrate 
and sulphate of aluminium. 
By the addition of a small amount of 
alum to water, it can be filtered 
through ordinary paper without diffi- 
culty, and yields a brilliantly clear fil- 
trate, in which there is no trace of sus- 
pended matter In our experiments, a 
solution of alum was added to the 
wrater, the whole well mixed by stirring 
or shaking, and then filtered after 
standing from one to fifteen minutes. 
So far as we are able to determine, the 
coagulative and precipitative action of 
the alum is immediate upon thorough 
mixture, and hence it is not necessary 
to allow the mixture to stand before fil- 
tration, but it can be filtered immedi- 
ately after mixing. 
To determine the amount of alum 
necessary to precipitate this water, 
alum was added in decreasing amounts 
to samples of water, which were then 
filtered through Schleicher and Schtill 
paper. In this way we found that the 
minimum limit was about 0 02 gramme 
of alum to one litre (1T6 grains to one 
gallon). Beyond that point the action 
of the alum began to be doubtful, and 
the water, although clarified by filtra- 
tion, was not wholly clear. To be sure 
of complete clarification, we took 
double this amount-0'04 gramme to 
one litre (2'3 grains to one gallon)-as a 
standard calculated to give certain re- 
sults'. This amount can be doubled or 
trebled without fear of any harmful re- 
sults, but there is no use of adding any 
more alum than is sufficient to do the 
work. The determination of the 
amount of solids removed from the 
water by the clarification with alum 
had not yet been finished. 
We consider it, then, as established 
that, by the addition of two grains of 
alum to the gallon of water, or half an 
ounce to the hundred gallons, water 
can be rendered capable of immediate 
clarification by filtration. The clear 
water obtained by filtration, after add- 
ing this amount ,of alum, contains no 
appreciable amount of free alum, and, 
in fact, in the majority of cases, ordin- 
ary tests fail to reveal its presence. 
While the clarification of water by. 
standing is very successful and well 
adapted to the treatment of large vol- 
umes of water, especially when time is 
not an element of importance, the case 
will very frequently occur that a rela- 
tively small amount of water is to be 
purified in a short time. In such a 
case not clarification alone is demand 
ed, but it is necessary that the opera- 
tion should take as short a time as pos- 
sible. Again, in order to make this 
method of clarification practical for 
domestic use, the operation of filtra- 
tion must be made extremely simple. 
No complicated or expensive apparatus 
should be used, but the filter must be 
made out of the simplest articles, such 
as can be found in every household. In 
this field there is an opportunity for the 
exercise of considerable mechanical 
ingenuity, and when the principles of 
the filtration are understood, and more 
is known about the different kinds of 
filtering materials, there will doubtless 
be many forms of house-filters devised 
' out of the odds and ends which may 
be at hand. 
It is not a difficult matter to get up 
a large filter that shall clarify many 
hundred gallons of water a day m 
an effective manner. Such apparatus 
already exists, and is used in manufac- 
turing establishments. In their con- 
struction, many points, such, for in- 
stance, as the cleansing of the filtering 
material, have been brought to a high 
grade of perfection. The difficulty 
lies in devising some form of simple 
and cheap filter which will filter a small 
amount of water as effectively as a rel- 
atively large amount, which will be 
always ready, will always work, will 
be so simple that any one can under- 
stand its operation, can be easily 
made, not easily broken, but easily re 
paired if broken, and which will not en- 
tail much extra work in order to get a 
clarified water. The filtering material 
must be cheap, easily obtainable, 
easily prepared, capable of being 
cleansed when clogged by use, or so 
cheap that it can be thrown away and 
replaced by new without appreciable 
expense. 
It is evident that the shape, size and ar- 
rangement of the filtering apparatus 
will depend very largely on the kind of 
filtering material used. Hence wTe 
began by experimenting on filtering 
media. The glass funnel and carefully- 
folded paper will be of but little service 
outside of the laboratory. But in cases 
of great importance, such as the prepar 
ation of water for the sick, this method 
is worthy of attention. 
In the large Hyatt filters a mixture of 
coarsely-ground coke and sand is used, 
and does most admirable and effective 
work. Granulated bone charcoal also 
makes a most excellent filtering bed. 
The most practical material for domes- 
tic use, however, so far as we have 
been able to ascertain, is cotton. Cot- 
ton batting can be bought in the shops 
for about ten cents a pound, and a 
pound of it will go a long way in filtering. 
It makes a coherent filtering layer, and 
when clogged by use can be cleansed 
by boiling up in wraterand rinsing, or, 
as it is so cheap, can perhaps as well be 
thrown away and replaced by new. 
The simplest form of filter for filtering- 
considerable amounts of water is a 
tube, one end of which is stuffed with 
cotton. A drain pipe is the best mate- 
rial, since it can be so easily cleansed. 
The plug of cotton should be from two 
to three inches thick, and may be held 
in place by a round piece of wood fit- 
ting into the bottom of the drain pipe 
at its shoulder, and secured by any 
suitable means. The piece of wood 
should be perforated, to allow the water 
to pass through. The shoulder of the 
pipe may be set in a circular channel 
cut in a piece of board, and by means 
of a central channel the water may be 
made to run off at a point of delivery. 
In our next report we shall present 
plans of simple filters, and the results 
of our experiments with them. 
The most practical form of filter for 
household use, and one that will easily 
filter a pitcherful of water in a short 
space of time, can be made out of a 4 
Purification of Drinking Water by Alum. 
bottle. The best form is the long kind 
in which sweet oil is sold, although al- 
most any kind of glass or earthenware 
bottle will answer. The bottom of the 
bottle is cracked off, and the sharp 
edge removed by rasping with a file. 
The cracking can be done by tying a 
thin, soft string, soaked in turpentine, 
around the place where it is intended 
to crack, leaving as small a knot as pos- 
sible, then setting fire to the turpen- 
tine, holding the bottle bottom up. 
Af ter allowing the oil to burn for an 
instant, the end of the bottle is placed 
quickly in cold water, when, if the 
operation has been rightly conducted, 
an even crack will be produced, and 
the bottom of the bottle will come off 
easily. 
A layer of cotton is now placed in 
the bottle. The cotton must be work- 
edin water, preferably warm water, in 
order to remove the adhering air, and to 
wet it well. A wad of the wet cotton is 
propped into the bottle and covers the 
mouth of the neck. Other pieces are 
dropped in, care being taken to build 
the layer up evenly, and to add the cot- 
ton in rather small pieces. After drop- 
ping them in, they should be pressed 
down and arranged by means of a rod. 
In this way a layer is made which 
should be from two to three inches 
thick. It should not be pressed down 
too tightly.else it may filter too slowly ; 
neither should it be too light, or water 
may form channels through it. After 
a little use the plug generally adapts 
itself. Particular care should be taken 
to be sure that the cotton is snug 
against both sides, since the water is 
liable to escape there. The plugs, how- 
ever, are easy to make, and a few 
attempts will soon teach one all the 
necessary manipulations. 
This bottle filter can be suspended 
or supported in any convenient way. 
Perhaps the simplest support is a 
block "of wood having an auger .hole 
bored through the centre, and the 
edges of the hole reamed out. In this 
hole the bottle sits securely, and the 
bevel of the hole catches the shoulder 
of the bottle, thus holding it upright. 
To use this filter, it is only necessary 
to pour the water, which has been pre- 
viously mixed with the right amount 
of alum, into it, when the clear water 
will run in a considerable stream from 
the bottom, and can be caught in any 
convenient receptacle. It is well to 
throw away the first tumblerful that 
runs through, if the plug is a new one, 
as a little sediment will pass through 
at first, but this soon stops. It is also 
advisable to keep the bottle nearly full 
while filtering, as this hastens filtra- 
tion. 
The mixing of the water with the 
alum previous to the filtration should 
be done in a separate receptacle. The 
only requisite here is that the vessel 
in which the mixing is done must be 
clean. A pail, jug, can, or any other 
vessel used in the kitchen will do. It 
is well to have the pail or can mark- 
ed on the inside with scratches so as to 
be able without difficulty to judge how 
much water there is in it, since the 
amount of alum should be added in 
about the right proportions. The eye 
gets very accurate in judging the vol- 
ume after a little practice, but it is 
better and just as easy to be accurate. 
A clean tin can of two to four gallons 
capacity is a good size, and, if possible, 
should not be used for any other pur- 
pose than for the drinking water It 
should be kept scrupulously clean, and 
after each use should be washed out 
and dried. It can be graduated by 
pouring into it a gallon of water, and 
marking with a file or other sharp 
point a scratch just at the level of the 
water. Then another gallon is poured 
in, and its level also marked. In this 
way a graduation is easily made which 
is sufficiently accurate for all the pur- 
poses here intended. As a rule, a can 
of four gallons capacity will be found 
quite large enough to filter the water 
used by a family of average size. The 
necessary amount of the alum solution 
is added to the water, the whole well 
mixed by stirring, and then poured 
into the filter. Here, again, one or 
two points should be observed. The 
mixing is best done with a long 
handled spoon. A very practical stir 
rer is a small cake turner, for by means 
of its fiat end a most thorough mixing 
can be effected. This mixer should 
not be used for any other purpose than 
to mix the water Experience shows 
that, if the vessels used for mixing or 
holding the water are not kept perfect- 
ly clean, the water may acquire a 
taste, and this will be laid to the pro 
cess instead of to lack of care. To fa- 
cilitate the pouring into the filter, it 
is well to have the can provided with a 
mouth or spout In fact, there is no 
form of can better than the regular gar 
den watering pot, with its long spout 
The solution of alum is made as fol- 
lows : Dissolve Imlf' an ounce of alum 
in a cup of boiling water, and when it 
is all dissolved, pour into a quart mea- 
sure and fill to a quart with cold 
water. (This solution should be kept 
in a bottle labelled ''Alum.") Fifty - 
four drops of this solution contain 2'3 
grains of alum, which is the amount to 
be added to one gallon of water. The 
old-fashioned teaspoon holds about 
forty drops ; the new spoons, however, 
hold about seventy drops. Hence, a 
modern teaspoon, scant full, will be 
about the right amount to add to every 
gallon of water to be filtered. No harm 
would be done if by mistake two tea- 
spoonfuls are added ; in fact, ten tea- 
spoonfu.s would have to be added to 
bring the amount of alum up to the 
fitrure recommended by Jennet (loc. 
c/Y.). A more satisfactory method will 
be to procure a small measuring glass. 
One fluid drachm will be the right 
amount. It will be found, without 
doubt, that the amount required for 
sr.me waters will beeven less than that 
suggested above. We would suggest, 
therefore, that those who use this 
method of clarification determine for 
themselves by experiment how little of 
the solution is required to make the 
water they use run through the filter 
perfectly bright and and clear.-Chem- 
ical Laboratory of Rutgers College.