Water purification

Document text

Google 



This is a digital copy of a book that was preserved for generations on Hbrary shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http : //books . google . com/| 



FROM 313 Wpist 102nd St., J^v, i,.. 



OOIViFL-IMELTjTS ok- 

Dr. G. G. Currier;, 



. \ 






• • • 

• • • • 



{^1 



[Reprinted from The Medical News, Nov. 6, 13, and 20, 1897.] 



WATER PURIFICATION H YGI EH I C ALLY CON- 
SIDERED. 

By C. G. currier, M.D., 

OP NEW YORK; 
ASSOCIATE OF THE AMERICAN SOCIETY OP CIVIL ENGINEERS, ETC. 

A FEW years ago the author, in various articles 
giving the results of experimental investigations on 
some of the subjects touched upon in these pages, felt 
bound to offer considerable evidence to prove that 
pure water is necessary to health, and that thousands 
of lives are yearly lost because of impure water.* 
Many epidemics due to a bad water supply have been 
reported at more or less length in the medical jour- 
nals and in the daily papers, which reach all classes. 
Thousands or even hundreds of thousands of deaths 
from cholera have occurred in Asia, in the Orient, 
in Egypt, and in Europe. The recent epidemic in 
Hamburg (1892) is still vividly impressed upon the 
minds of those who give any thought to the question 
of water-borne diseases. 

Fortunately the broad oceans safeguard us to a 
large extent against cholera and other Asiatic plagues; 
but here, as well as elsewhere, typhoid fever, not to 
speak of various lesser diarrheal disorders which are 
likewise communicated by impure drinking-waters, 
yearly causes thousands of deaths and many more 
cases of disease of varying degrees of severity. Even 
malaria is regarded by numerous intelligent physi- 
cians and other acute observers as presumably a 

1 During the last few years there has been furnished so much 
statistic and other testimony to this effect that it would require 
too much space to reproduce it here. 



fi^^t^ 



water-borne disease, although the evidence offered 
does not suffice to prove that contention. Yet 
typhoid fever alone is a sufficiently grave peril 
to warrant an immense amount of agitation and 
the expenditure of many millions of money to pre- 
vent its occurrence. It is very well known that 
when the water-supply of any isolated household 
or of any community is naturally pure or adequately 
purified by artificial means, typhoid fever does not 
occur unless water be casually consumed from an 
impure source, or unless milk or other food, used 
in an uncooked condition, have come in contact with 
infected water, or unless the victims have acquired the 
disease directly from someone suffering from it. For 
all industrial and domestic purposes it is desirable that 
water be as pure as possible. Whether for * ' feeding * * 
a boiler or quenching thirst, for watering the streets 
or cleansing a wound, the purest water is the most 
satisfactory. The enlightened economist and the 
sanitarian insist upon this, and, even if space per- 
mitted, a citation of the volumes of facts and figures 
available to prove it would not be necessary for those 
familiar with the rudiments of modern sanitation. 

What then is pure water, hygienically speaking? 
Briefly, it may be said to be clear, palatable, free 
from bacteria in large numbers, and wholly free from 
bacteria which can cause disease; it must contain no 
lead or other poisons, and have only a permissibly 
small amount of mineral salts. The salts of lime and 
magnesia, in a supply for general industrial and domes- 
tic use, must not be much in excess of 15 parts in 100,- 
000 of water, otherwise it is too hard. The water 
of the best lakes and rivers may be less than one- half 



or even one-fourth as "hard ' * as this. It should bd 
mentioned that the salts present in the water which 
we drink do not count for much in the total amount 
needed daily by the human body (nearly two ounces 
of salts being required), as that amount is abun- 
dantly supplied by a mixed diet. The daily amount 
of water swallowed by a healthy man averages about 
a gallon. This quantity of good natural water usually 
does not contain more than ten grains of salts, and a 
mineral water having one hundred grains of salts to 
the gallon is altogether too strong for large amounts 
of it to be tolerated. A slight proportion of lime is 
hygienically desirable in water-supplies, for it restricts 
the tendency of the water to act upon metals. A 
very soft water more readily takes up lead, for in- 
stance, from lead pipes in which it is conveyed. A 
very soft moorland water, especially if acid and well 
oxygenated, is more apt to cause lead- poisoning to 
develop than is the case with limy waters. 

. Clearness, to the uninstructed, seems to be the 
most important quality of a water, and yet the clear- 
est water, particularly that of wells, often causes 
t)rphoid fever, for the causative bacteria of that dis- 
ease are not detectable by the senses when they have 
been introduced into a supply. On the other hand, 
turbidity is not usually due to a dangerous cause, 
however unpleasant it may appear to the eye. Water 
may be perfectly wholesome, and yet so colored by 
harmless vegetable matter as to arouse distrust among 
those unfamiliar with it. C hemic tests of water are 
of most importance nowadays as aiding one to deter- 
mine its mineral constituents. In studying a pro- 
posed new supply, especially if it be from a peaty 



source, it is well to know its degree of solvent action 
upon lead, for it is quite expensive to construct and 
maintain the plant needed for counteracting such a 
tendency, by filtration through lime or by the suitable 
addition of lime or soda carbonates. Of the organic 
constituents, an undue quantity (more than .015 in 
100,000 parts) of albuminoid ammonia is the impor- 
tant practical factor which aids us to determine 
whether a water is bad or not. Wankl)m, in his 
latest work on ' * Water Analysis, ' ' does not attach 
much importance to the presence of nitrates, although 
they are usually regarded as evidence against the 
purity of a water. The elaborate organic chemic 
testing of a sample of water is not of nearly as much 
hygienic value as a careful local inspection by a 
skilled hydrologist of the sources whence the water 
comes and a critical searching for possibilities of 
present or future contamination. 

Chemistry, then, being an inadequate guide, ex- 
cept as regards the presence of metals and salts, can 
any degree of reliance be placed upon the revelations 
of the microscope as to whether a water contains or 
is likely to contain matters or organisms prejudicial 
to health ? Not to any sufficient extent. Of course, 
if tapeworm eggs or partially digested meat fibers be 
detected in a sample of water, it evidently is pol- 
luted, and must be condemned; but it is hardly 
probable that such things, even if present, will be 
found to have come into the exceedingly small field 
of a microscope. Nor does the recognition of pro- 
tozoa in water aid us hygienically, according to 
Neisser, who has made a special study of this sub- 
ject. While a water rich in organic matter is likely 



s 

to contain many micro-organisms, it may not have 
many protozoa. On the other hand, both bacteria 
and protozoa can thrive in water that is very pure 
indeed. Wash -water, sewage, privy-vaults, and ma- 
nure-heaps are rather lacking in protozoa, although 
affording most unwholesome contaminations of drink- 
ing-water. Every leaf or blade of grass that falls 
into a well introduces more protozoa {monadineSy 
amebcBy ciliata) than would human feces. Protozoa 
abound in hydrant water at all seasons. 

During the last fifteen years an incontrovertible 
mass of evidence has accumulated to prove that when 
drinking-water causes cholera, typhoid fever, and 
other diseases it is because of the presence in the 
water of the bacteria peculiar to that disease. While 
heretofore none of the various methods in use have 
enabled us to always recognize with certainty the 
disease- producing bacteria which are present among 
the numerous microbes of harmless character always 
to be found in ordinary water, it seems as though we 
are on the threshold of important discoveries in this 
direction. Even the familiar gelatin- culture method 
is of considerable value in the service of an expert, 
although local inspection is always of paramount 
importance. 

The usual way to make the gelatin test is to add 
some of the water to strong, specially prepared, 
sterile bouillon made solid by the addition of gela- 
tin, as explained in The Medical News by the 
writer (April 27, 1889). This is done carefully after 
the gelatin is softened by gentle heat. Then the 
gelatin is allowed to ''set," and in two days, more 
or less, the isolated bacteria of the original drop or 



more of water added, being diffused evenly through- 
out the mass of nutrient gelatin, will have increased 
enormously in that favorable medium. Accordingly, 
each separate one of the bacteria (xiriirTr ^^ ^^ ^^^^ 
in size), although of itself visible only by the aid of 
high powers of the microscope, has, by multiplying 
to millions, formed a visible dot, cup, or other pe- 
culiar appearance separate from the others. 

Without taking the necessary space to explain the 
many details of bacteriologic work which do not bear 
directly upon the subject of this- paper, it may be 
briefly stated that, after years of discussion of the 
value of the mere counting of the number of bacteria 
present in a cubic centimeter of a carefully taken 
fresh sample of water, as a rule a water which has 
few bacteria is better than one which has many. The 
best waters have usually less than one hundred bac- 
teria in each cubic centimeter (nearly one-fourth ot 
a teaspoonful) . Yet a large number of the organisms 
should'not necessarily be considered as indicating 
that the water is unfit for use; and, on the other 
hand, disease bacteria entering a water which normally 
has few bacteria may produce serious infection. Of 
course, the species of bacteria present must be con- 
sidered in forming an opinion. Eisner's method, 
consisting in using * ' potato gelatin ' ' containing one 
per cent, of iodid of potash, promises to aid materi- 
ally in the recognition of typhoid germs in water. 
For testing the efficacy of filters or other means of 
rendering an infected water fit to drink, the gelatin- 
culture test is all-important. It is the bacteria in 
such a water that the sanitarian regards as the ele- 
ment of danger. If the process removes all of these. 



it effectually purifies the water; and if it removes 
or destroys all but a minute proportion of these 
micro-organisms, it is safe to assume that practically 
all of the disease-producing germs, if any were pres- 
ent, have been removed. That the water produced 
is clear or palatable is, of course, readily determined 
by the senses. 

In testing a filter in operation, the original water 
is examined from time to time by collecting average 
samples in heat-sterilized bottles. A cubic centi- 
meter, or half as much, of a sample is taken in a 
sterile glass pipette and added to a cotton -plugged 
test-tube containing ten cubic centimeters of nutrient 
gelatin which has been softened by gentle warmth. 
After these are commingled the gelatin is poured 
upon a cool plate of sterile glass or a fiat glass dish, 
and this is then promptly covered and labeled. An- 
other test-tube of gelatin is treated in the same way, 
but is plugged with its sterile cotton as soon as the 
water is added. After mixing this in by gentle agi- 
tation, it is marked and laid upon its side so that the 
fluid gelatin and water reaches almost up to the cot- 
ton. Frequent samples of the filtered water are 
taken with care to prevent the entrance of bacteria 
from without. These are plated and tubed in the 
way just indicated, and all the plates and tubes are 
then kept away from dust and strong light, under 
identical conditions of free air, light, and moisture. 
Those in which the water introduced was sterile will 
show the gelatin unchanged even after a week, while 
those into which bacteria entered with the water will 
show a varying number of colonies corresponding to 
the number of bacteria entering. Usually those from 



8 

filter-samples are of fewer varieties, even if — as in 
the case of low-grade or badly cared-for filters — 
greater in number than with the same amount of 
water from the raw supply. 

Natural purification by slow flow through many 
feet of fine, porous earth produces water absolutely 
free from harmful or other bacteria which may have 
been in the water or on the surface where it fell. 
The majority of investigators find that, while bac- 
teria abound on and very near the surface, they are 
greatly lessened a yard beneath the surface of the 
ground, and five or six feet deeper, in fine soil, none 
of these have penetrated. Accordingly, such ground- 
water, whether issuing as uncontaminated natural 
springs, or even obtained by properly constructed 
and protected wells, is very pure as regards bacteria, 
although it may have taken up a considerable amount 
of mineral matter. Water evaporated from oceans, 
lakes, and the top of the ground, and precipitated as 
rain, is then, like distilled water, very soft and pure, 
although that which first falls takes up some impuri- 
ties from the air, especially near cities and from the 
roofs from which it is collected. In very many exami- 
nations made in various parts of this country the 
author, like others, has regularly found that brook- 
lets in unsettled regions have very few bacteria, and 
that larger streams and lakes supplied by springs and 
rainfall from watersheds which are uninhabited and 
untilled have, as a rule, few bacteria. But as human 
habitations and industries increase, the number of 
bacteria present also increases, and a river in an ag- 
ricultural valley, even without sewage entering the 
stream, has many more bacteria than a similar one 



in an uncultivated district under the same conditions 
of rainfall. 

Communities which derive an abundant supply 
from the best of such upland sources, and which can 
preserve the watershed free from all permanent or tem- 
porary human contamination, have an ideal water- 
supply. Unfortunately, such pure supplies are alto- 
gether too few and do not increase, while the de- 
mand for water increases constantly, and must be 
met. Furthermore, supplies which were originally 
pure become contaminated. Thus, the necessity 
arises of purifying an inferior water so as to make it 
healthful and yet have an abundance for domestic 
and industrial purposes. The masses cannot or will 
not afford the outlay and trouble involved in using 
bottled waters when the public supply is bad. Neither 
is it found practicable to have a separate system of 
hydrants and pipes for a scanty pure supply when 
the main supply is unfit to drink. 

If a water be regularly turbid and contains visible 
dirt, a general outcry for some process of treatment 
is much more likely to arise than when it is clear and 
palatable, yet containing the germs of infectious dis- 
eases. It is in the latter case that sanitarians find 
fault with the water. When turbid and colored with 
dirt and vegetable matter, there will occur numerous 
complaints from the users. If inspection of the water- 
shed and the evidence of skilful bacteriologic exam- 
inations make it seem likely that dangerous contam- 
ination exists and cannot be checked or prevented in 
the future, theA^it is well to have recourse to filtration 
on a large scale. This renders the water more agree- 
able to the eye and the palate, while the process 



lO 



properly carried out removes nearly all the bacteria. 
It is not easy to determine just how much danger- 
ous contamination may be tolerated in a water with- 
out manifesting itself in the form of disease and death 
among those using the unpurified raw water. When 
a supply is condemned because obviously contam- 
inated, the argument may be brought forward that 
the germs of cholera, t)rphoid, and other diseases 
tend to die out soon after entering a stream or body 
of water, and that that water purifies itself. It is 
true that something of the sort takes place. The 
writer studied this question * and found that dilution 
and sedimentation, the influence of light and oxygen 
(of the air), and also the antagonistic action of vola- 
tile acids in some river, and of the hardy water-bac- 
teria present in all waters, especially in impure ones, 
contribute to the purification after contamination by 
the entrance of harmfiil bacteria with sewage or other 
filth. Yet if Eisner's test is employed, it is found 
that typhoid germs survive much longer and are 
much more prevalent than was formerly thought.* 
Accordingly, and in view of the seemingly small 
amount of infection causing some very serious epi- 
demics, it is wisest in case of doubt to regard all 
filth contamination of water as dangerous. It is not 
safe to assert that all rivers of this country can 
*' purify themselves " within twelve miles, as is often 
stated. Each case must be considered by itself, and 
judged independently. Yet it is difficult to convince 
municipal authorities of this necessity for pure water. 



> Trans. Amer. Soc.of Civ. Eng.^ February, 1891, and American 
Journal of the Medical Sciences^ December, 1890. 
' New York Medical Journal^ April 10, 1897. 



It 

and some of our cities have water supplies which may 
be classed as dilute sewage. As a consequence, 
thousands of cases of typhoid fever result among those 
drinking such impure water, and the disease is much 
more comtnon among us than in the well-managed 
cities of Europe. 

Recognizing to a certain extent the possible dan- 
ger in drinking impure water, but also being influenced 
chiefly by its unpalat ability or unpleasant appearance, 
intelligent people try to do something individually 
to purify the small amount used by their households. 
Straining the water through clean flannel or other 
dense substance, they find it clearer. If boiled, in 
addition, it is also free from any element of danger 
to those who drink it. And yet, prepared in this 
way, it has an unpleasant taste for one not used to 
it. This is because the two per cent., or less, of air 
in the natural water has been driven off by the boil- 
ing. If we restore this by forcing air through it by 
a bellows or by pouring the water several times from 
one pitcher to another, the taste is improved. The 
addition of a little cold tea, lemon juice, seltzer 
water, or other chosen savor, disguises the taste. 
But boiling, though tlje most generally available and 
reliable means of artificially purifying water, requires 
effort, especially when bottles or similar containers 
are used for holding the sterilized water, since they 
have to be kept clean. If the water be cooked in 
the bottles, this involves some breakage. When, 
therefore, a housekeeper is offered '* pure drinking- 
water guaranteed, * * for an outlay of 1 7 cents, which 
purchases a " self-cleaning water-filter ' * which has 
embossed in large letters upon its nickeled metal 



12 



Case the pretentious words, * 'germ -proof filter,** the 
temptation is strong to employ such a means of im- 
proving the domestic supply. These quoted words 
are from the printed advertisement of one of the 
large department -stores of this city. The filter, un- 
fortunately, does not purify the water as claimed, 
although it strains out coarse dirt and makes dirty 
water somewhat clearer. It is as valuable as any of 
the ordinary filters, and is perhaps least likely of all 
of these to become a source of infection by serving 
as a breeding -ground for bacteria. It is filled with 
crushed quartz held in place by two wire screens. 
Screwed upon a faucet, when clean, it allows a con- 
siderable flow of water. Soon, however, the flow is 
obstructed by the accumulated dirt. If then, such a 
filter be reversed and the other end be screwed upon 
the faucet, the coarser dirt is washed out and also 
much of the finer. Yet considerable fine dirt and, 
among other things, many bacteria remain attached 
to the sand. Boiling kills these, but if this be 
omitted, and the filter allowed to remain unused for 
some hours, as over night, at ordinary room temper- 
atures, the adherent bacteria will have increased by 
multiplication in the organic matter present. Ac- 
cordingly, the first flow of water from the faucet after 
such disuse of the filter will contain many more bac- 
teria if it flow through the filter than if the filter were 
not used. The number lessens gradually as the water 
continues running. 

This serves as a type of all ordinary domestic fil- 
ters, in which class is included for brevity all, 
whether straining water from a faucet or from a jar, or 
other receptacle under gentle pressure, and in which 



t3 

the straining substance is not very dense. In 
The Medical News for April 27, 1889, the writer 
reported the results of prolonged experiments upon 
very many filters made of the various substances used 
for the purpose, and since that time he has tested 
every kind on the market. Nearly every permeable 
substance has been called into requisition by in- 
ventors in this line, such as cotton felt, wood felt, 
wood pulp, wood, cloth, sponge paper, chaff, wire 
gauze, charcoal, carbon, polarite, spongy iron, por- 
ous porcelain, porous stone, and mineral substances 
of varying composition in a natural or prepared 
state, asbestos, and simple sand. As a rule, these 
are the products of wholly untrained minds quite un- 
familiar with the principles underlying successful 
filtration. Now and then there comes a new varia- 
tion upon, or an independent application, of some 
previous idea. The study above indicated, showed 
that except in so far as it is intended to be a mere 
strainer of more or less of the visible dirt, the essen- 
tial virtue of a filtering substance is fineness of struc- 
ture such as will hold back all bacteria under ordinary 
conditions of usage for a day or more. If not of such 
dense material, it is merely of service as a strainer 
through which bacteria can pass, however well the 
water is cleared up, and in which they can increase 
to such an extent that the water flowing from them 
has more bacteria than that which they are supposed 
to purify. 

This element of bacterial contamination from the 
filter itself varies with the kind used, and of course 
disappears for a brief time after the filter is sterilized, 
as by boiling. Accordingly, such imperfect filters 



should be simple in construction, commodious, easy 
to cleanse, and not destructible by frequent boil- 
ing. As to their capacity for holding back bac- 
teria, that varies with the density of the filtering 
substance and its strength and durability. If it tears 
readily and allows holes or fissures to form, bacteria 
can of course pass through such breaches of continu- 
ity. The simplest and most permeable kind — ^the 
sand filter or, rather, strainer above described — holds 
back so few bacteria as to be practically worthless in 
this respect. A dense, freshly sterilized flannel bag 
upon a faucet or a folded filter paper (or even a plug 
of absorbent cotton) in a funnel does rather better 
work. Carbon blocks and wood pulp hold back a 
considerably larger percentage, but allow disease bac- 
teria to pass through and also to increase in the por- 
ous filtering mass. Porous stone is still more efficient 
and when of the finest quality will hold back larger 
and medium-sized bacteria, such as the germs of 
cholera and typhoid fever, for many hours. It ren- 
ders some turbid, silty waters very clear. While the 
best of these, frequently sterilized by chemicals or by 
boiling, can be employed by an expert with safety, 
they are not to be recommended for ordinary use in 
time of epidemics or when the water filtered is regu- 
larly infected, unless it has always been cleaned and 
boiled a few hours after it has first been used after a 
previous boiling. This is a requirement which the 
masses will not observe. So long as a filter renders 
water clearer, they allow it to strain their drinking- 
water, regardless of the possibility of bacteria pass- 
ing through. Asbestos, in carefully arranged sheets, 
is employed for filters, some of which display great 



IS 

mechanical excellence. This can be made to yield 
better results than any ordinary filter, although in- 
ferior to the best; but the weakness of the tissue, the 
bulkiness or complexity of the efficient filters of this 
substance, and the difficulty of cleansing them, mili- 
tate against their adoption, although some may be 
sterilized by baking. Porous porcelain of coarse 
texture, such as is used in electric-battery cups, is 
much like porous stone, holding back bacteria for a 
brief time, but is quite inferior to finely porous por- 
celain as used in the most efficient filters. 

These second-rate substances are bacterially effi- 
cient, in proportion to the minuteness of their largest 
pores and in inverse ratio to the pressure with which 
the water is supplied. The densest offer the greatest 
resistance to bacteria and also to water. A corollary 
of this is the fact that a thoroughly efficient filter 
cannot be very rapid, and when a given filter yields 
water much more rapidly than another having the 
same surface exposed to the crude water, the former 
is less capable of separating bacteria. Both being 
under the same conditions, the less rapid yields water 
freer of germs than the other. And yet the slowest 
filters are not necessarily in all cases the most effi- 
cient, as when, for instance, the slowness is due to a 
thick mass of fibers or granules which cannot alto- 
gether prevent bacteria from getting through, and 
which, in consequence, must allow them to increase 
prodigiously in the dirty mass. In no case is slow- 
ness a merit, but, because of exhausting the patience 
of those who use a filter, is to be lessened as much 
as is consistent with practical retention of " germ- 
proof quality. 



i6 



The only filtering substances which can safely be 
recommended for household or other use on a small 
scale, is (i) fine, dense, porous porcelain of standard 
make and uniform quality tested by the maker; and 
(2) the prepared infusorial earth cylinders which 
carry the trade name of Berkefeld. The latter re- 
main germ-proof for a shorter time than the best of 
the former class — of which the ''Pasteur" filter 
cylinders may be taken as a type (Fig. i), yet they 

Fig. I. 




Pasteur-Chamberland filter tubes. Portion of wall of upper one 
removed : 

are within the limit of safety if cleaned and sterilized 
(by boiling) daily. They have the great merit of 
comparatively rapid flow, and are less likely to be- 
come broken than the French tubes. Both of the 
two specified are expensive, being patented articles 
of foreign manufacture. There are several patterns 
of filters using excellent porcelain material but made 
in this country at Trenton and elsewhere. Specimens 
tested by the writer were quite as good in germ- 
proof quality and flow as the material of the Pasteur 
filters. This he has proved by the most rigorous 
tests carried on in one instance for eight weeks of 
unintermitting flow, during which time no bacteria 
came through. 

The '' Pasteur " filter is also, and more properly, 
designated by the name of Charaberland, who first 



17 

exploited this valuable material, taking out a Belgian 
patent (No. 69,095) for the exclusive right of using 
rebaked porous porcelain for filtration. A German 
patent, antedating this, and a French one dated 
eighteen months later, might be interpreted as in- 
dicating that Chamberland attached importance to 
the principle of filling with charcoal the space within 
the metallic casing outside of the porous cylinder. 
This idea was not utilized by the producers, although 
in a less popular competitor (Malli^ filter, patented in 
February, 1895) are used flat plates of similar porous 
porcelain outside of which are corresponding plates 
of carbon to intercept dirt and prevent speedy clog- 
ging of the porcelain by turbid water. Before know- 
ing of Mr. Chamberland' s idea, the writer had sug- 
gested the same expedient for use with muddy water. 
For the Chamberland -Pasteur tubes, semivitrified 
porcelain is ground up and used in the preparation 
of the finest product, to give greater porosity. Finely 
ground silica answers the same purpose, yet if any but 
the smallest sand-grains are used, it is considered 
by the makers that the resistance to bacteria is likely 
to prove insufficient. Some of the best porous por- 
celain produced in this country seems to be composed 
of from fifty- five to seventy per cent, of finely ground 
silex, with between forty-five and seventy per cent, 
of kaolin clay. This porcelain is quite inexpensive, 
fairly durable, and as germ -proof as the best French 
product. The German patent. No. 19,589 (January, 
1 882), covers the preparation of a similar porous filter- 
ing material, as is also the case with the English pat- 
ents of Bagshaw (1834), Roberts (1859), and Joseph 
(1870). Yet it is doubtful whether the products were 



i8 

fine enough to be germ-proof. It was not until 1871 
that it was definitely recognized that porous porcelain 
of the proper sort could remove all bacteria from 
liquids. This was discovered and utilized by Tiegel, 
a Swiss pupil of Klebs. About the same time Zahn 
employed clay cells for the same purpose. In the 
same way Hiller (1874 or earlier), Panum, and others, 
as well as Sanderson (1878), utilized porous porce- 
lain. In France, Arloing, Toussaint, and others 
achieved the same result as Tiegel by using plaster 
with or without the admixture of asbestos for strength- 
ening the texture. Pasteur, in 1877, used this also, 
abandoning it later for rebaked porcelain. The 
admirably adapted cylinder shape still used in the 
Pasteur filter is due to Chamberland, although some- 
what similar patterns had been illustrated in publica- 
tions previous to his time. The efficacy of rebaked 
porcelain was announced by him in 1884, through 
Bouley.* While formerly made of two degrees of 
porosity, the output sold here is now of only one 
uniform density. 

This substance, thus gradually evolved, has never 
been improved upon as regardspermanency of germ- 
proof quality. A perfect cylinder will remain im- 
permeable to bacteria for days and even weeks, al- 
lowing water to flow through in a slow stream or in 
drops, until the silt and other things carried by the 
crude water choke up the interstices on the surface. 
This extreme slowness, increasing with use, is a 
serious drawback, which, like the brittleness, renders 
it less popular than its admirable filtering property 
would otherwise make it. A cracked or flawy cylin- 

* Comptes Rendus, Acad, des Sciences^ August 4, p. 247. 



19 

der should be discarded, like a broken one. The 
frequency of mishaps with ordinary careless handling 
and cooking for the purpose of sterilizing make the 
Pasteur filter too expensive for the masses, since a 
new cylinder is retailed for three or more times 
as much here as in France. There the writer has 
bought them for about thirty cents. A cylinder 
which has deteriorated by weeks or months of use, 
but which is not cracked, can be renovated; yet the 
ordinary cook cannot be trusted to do this. Care- 
ful baking of cleaned and dried cylinders at the high 
temperature of nearly 300*^ C. (572° F.) will make 
the porcelain nearly as efficient as when new. This 
heat, of course, completely sterilizes the cylinders at 
the same time. They should be allowed to cool 
gradually. Another method is to use chemicals. 
The tubes are well cleaned and then inverted to 
drain out tor fifteen minutes or more. Then, out- 
let upward, they are filled with, and soaked for at least 
fifteen minutes in, a twenty- per-cent. lukewarm solu- 
tion of pure chlorid of lime (or Labarraque's solution 
or eau de Javelle with twice its volume of water). 
Thereby the albuminoid and mineral deposits in the 
pores are disorganized. Next they are emptied and 
then filled with, and lie for another quarter of an 
hour in, hydrochloric acid diluted with four or five 
times its volume of water. After this they are emp- 
tied and rinsed with boiled water. Another means 
of sterilizing and regenerating them is to employ 
similarly, and for twice as long a time, a 1-150 so- 
lution of permanganate of potash, followed by a five- 
per-cent. solution of bisulphite of soda. Very dilute 
hydrochloric acid tends to prevent the deposition of 



2C 

manganic bioxid in the pores. To detect flaws or 
other imperfections, the manufacturers of good por- 
celain filters moisten the substance and then apply 
air-pressure of at least seven pounds. Those which 
allow air to pass are rejected. The test is very easy 
for any one to make by means of a small air-pump. 

After a few days, or at stny rate a week of use, such 
a filter-cylinder needs to be removed, brushed off, 
and rinsed. Then it should be boiled gently for 
half an hour and allowed to cool gradually before 
replacing in its casing. This is because not even 
the best filter-substance is germ-proof for an indefi- 
nite length of time. Until nearly a decade ago they 
were regarded as absolutely germ -proof, and even 
later than that some observers, presumably after in- 
adequate tests, have said that these porous porcelain 
surfaces allowed no bacteria to get through. In 
1888, by a series of thorough tests and checks 
thereon, made at the Carnegie Laboratory in New 
York, the author found that, even when the filters 
were kept cool and when only cold water was used, 
bacteria passed through soon after the end of the 
second week of constant usage; so that in time many 
more were in the filtered than in the unfiltered water. 
Pasteur cylinders used for the purpose were of the 
pattern sold at present, and of the best quality, hav- 
ing been purchased by the writer in Paris, of the 
principal agency, and represented by the seller as 
perfect. The results of those experiments were ac- 
cepted, and have since been confirmed by other in- 
vestigators. Disease- bacteria can grow through this 
substance in laboratory experiments where every- 
thing is arranged to favor the growth; but it has not 



yet been shown that this can occur under ordinary 
household conditions of usage of tightly fixed, per- 
fect cylinders washed and sterilized by boiling at 
least once every week. 

These filters possess, then, a high degree of effi- 
ciency as regards the removal of bacteria, and what- 
ever has been said above regarding the efficacy of 
Fig. 9. 



Elattery of filters of Ihe BerkefeLd paUcrn. a, inflow; *, outflow, 

one well-known pattern applies equally to all filters 
properly constructed of the best porous porcelain. 
As already stated, American porcelain stands the 
severest bacteriologic test as well as does the im- 
ported material. There seems no inherent reason 
why all of these should not be fairly good if only 
made with proper care. It must be borne in mind, 
however, that when of small surface they cannot be 
rapid and at the same time germ-proof for any con- 
siderable length of time. 



22 



When ' ' batteries ' ' are arranged by connecting a 
few or many of the separate porcelain tubes, of the 
Pasteur type, or of similar styles, to an outflow pipe 
by very tight joints, correspondingly increased 
quantities of pure drinking-water can be obtained 
from them. These multiple arrangements do not 
prove quite so ''germ-proof" as the simple tubes 
securely fixed each in its separate metal casing, yet 
in an institution where a physician, or other person, 
trained in simple bacterial methods, constantly su- 
pervises the arrangement, the cleansing and neces- 
sary renewals, and also regularly tests the yield, very 
bad water can be rendered wholesome by such 
means. This is proved from the reports of the French 
Minister of War, based on the prolonged and sys- 
tematic use of these filters in French garrisons, result- 
ing in a reduction of the typhoid death-rate from 
one to ten per cent, of what it formerly was, and the 
cases occurring are said to be nearly always explaina- 
ble by the use of the ordinary or some outside supply 
in an unfiltered state. As a point in favor of these 
and against inferior filters, it is stated that in the 
British army in India, where charcoal and sand-fil- 
ters have been used, the number of deaths from ty- 
phoid fever has steadily increased since 1884. Such 
filters are worse than none. Outside of garrisons, 
military use of porcelain filters proves, with or with- 
out the Andr^ cleaner and infiisorial earth powder, 
to be unsati factory, notably in campaigns where the 
foulness and pestilential quality of the only waters ob- 
tainable make an adequate purifier absolutely neces- 
sary. However brilliant the laboratory and exposi- 
tion results, the military filters failed under the 



23 

conditions of practical usage. Barth^l^my, and also 
Molinier,^ very forcibly pictured the inadequacy of 
these. The best pattern among military filters is 
called ^'filter de campagney' has fifteen filtering cyl- 
inders, and weighs no pounds or more. 

For use in households where no hydrant water or 
other piped supply exists, yet where filtration of the 
drinking water seems necessary, the writer would 
suggest that some manufacturer introduce circular 
plates of this germ -proof porous porcelain, perhaps 
slightly convex on the lower surface, in order that 
the water might drop only from the center of the 
plate. This could be held to the necessary cylinder 
or section of pipe above, and to which it will form 
a bottom, by screw-clamps pressing it tightly against 
an intervening rubber ring on the periphery. Al- 
though such a plate would filter very slowly, it could 
be very easily cleaned and boiled. Porous stone ot 
similar pattern is used, but is unsatisfactory, not be- 
ing germ-proof, although it clears turbid water. 
Being cemented into the lower rim of the jar above, 
it cannot well be sterilized. There are various pat- 
terns of non -pressure filter metal jars using several 
porcelain tubes inserted in holes in the bottom. 
Siphoning arrangements may also be employed. The 
yield is often contaminated. 

The porcelain tube-filters, as also those of the ma- 
terial about to be mentioned, may be used with a 
union-joint or two-way water-cock. The cumbrously 
long Pasteur filter is thus rendered more convenient 
for ordinary sinks, yet the writer prefers the older 
patterns in which the water flows directly downward 

> Archives de Midiciney Navale et Colomale, 1893 and 1894. 



24 

out of the porcelain cylinder into jars or other recep- 
tacles, even though thaX necessitates a special faucet 
reserved for the use of this filter. In the case of a 
specially introduced piping the water should come 
through iron instead of lead, or a block-tin pipe could 
be used at twice the price of lead pipe. 

A successful rival to the above filter is presented 

by the (Nordtmeyer) Berkefeld pattern which is 

Fig. 3. 



Berkefeld filter. 

similar in shape and mechanism, but which is many 
times faster, yielding when clean and freshly boiled 
nearlya quart of germ-free water in a minute from the 
smallest domestic model. The porous filtering cylin- 
der is only three inches in length. Closed, like the 
porcelain cylinders, at one end, the open end is firmly 
cemented into a flattened, metal funnel having a 
small central outflow tube, by the outside of which it 
isscrewed tightly inside its metal casing. {See Fig. 
3.) No breakable part is exposed when the fil- 



25 

ter is attached to a faucet. The household pattern, 
illustrated here, is so compact that it can be attached 
to a kitchen faucet and yet allow a jar or pitcher to 
be placed beneath to catch the filtered water. Its 
comparatively rapid yield is due to its extreme por- 
osity. The short porous cylinders or cups have 
walls more than half an inch thick, and are com- 
posed chiefly of infusorial earth appearing under the 
microscope as compressed diatoms (fossil bone-meal) 
held together by a secret process. They cost $1.50 
each, and are more durable than porcelain. The 
general rule above given, that with increased yield 
the germ- proof quality is lessened, holds good with 
these as with other filters, yet they hold back bac- 
teria long enough to render them safe if sterilized 
by boiling every day or two. The writer, who has 
tested a number of these during the past two years, 
has not found one that could not hold back all bac- 
teria for one day. In some experiments, under vari- 
ous conditions of practical usage and with a pressure 
as high as thirty- eight pounds, when the filters were 
kept cool (io°-i2° C.) during an intermittent yield 
of more than five gallons daily, they produced sterile 
water until the eighth day. 

From the results of many tests, it may be stated 
briefly that the Berkefeld filter will hold back all bac- 
teria from one to three days. We can safely recom- 
mend them, therefore, even for use with infected 
waters, on the condition that the porous filtering 
cup be taken out daily, and cleaned and then steril- 
ized by boiling for half an hour. The water should 
not be very warm when they are put into the clean 
vessel in which they are to be boiled and should be 



26 

allowed to cool enough to be endurable by the hand 
when they are taken out. When greatly thinned by 
repeated brushing or whenever a flaw develops in a 
j)orous cup, it should be discarded. It is advisable 
to have an extra tube on hand, although a single 
tube can be used for many months. The filtering 
substance lasts longest if not brushed roughly, but 
rather cleansed by means of pieces of cloth under a 
flow of water upon the cup. They do not break by 
gentle boiling, with little or no ebullition, even if 
for convenience they be thus sterilized in the same 
clean vessel in which other things — such as nursing - 
bottles — are cooked; yet rough handling breaks them. 
As with porcelain tubes of different makes, when the 
water is not supplied under pressure, these can be 
arranged to work with various forms of portable or 
fixed pumps, singly or in batteries. A single tube 
thus arranged weighs, with pump and portable stand, 
eleven pounds, and yields a gallon in two minutes. 
At least one extra tube should be at hand and also a 
vessel for the daily boiling, which must not be 
omitted. For bottling establishments and other 
users of large amounts of clear waters, batteries of 
these tubes are successfully used. One of fifteen 
tubes freshly cleaned and boiled yields seven gallons 
of sterile water per minute under a hydrant pressure 
of forty pounds. Testing such a filter under a lower 
pressure (East 426. St., New York), the writer found 
the water quite free from bacteria, the very few that 
were present probably coming from the unsterilized 
hose pipe used for conducting the outflow. The 
water was as pure as that from any mechanical filter 
which the writer has examined. Yet stress must al- 



27 

ways be laid upon the importance of frequent boiling 
of such tubes. Reliance must not be put upon the 
employment of automatic means of cleansing them 
by the use of powdered infusorial earth to the neglect 
of boiling several times a week, although bottlers are 
apt to disregard such injunctions. In a carefully 
managed hospital or institution under proper scien- 
tific medical control these battery filters can produce 
water entirely free from germs. 

All things considered, the simple form of the 
short- cylindered Berkefeld filter is to be recommended 
as the best for household use when there is a hydrant 
supply needing filtration, but it must be repeatedly 
emphasized that frequent boiling of the porous cup 
is essential to perfect results. While thus commend- 
ing this filter, the writer would not disparage the 
best porcelain filters. In his own residence, both 
kinds are in regular use and both yield sterile, whole- 
some water. 

As compared with the use of these germ-proof fil- 
ters by the few intelligent people who will regularly 
give them the requisite amount of care (in itself very 
trivial, but yet indispensable) any other filtration of 
drinking-water is inferior. Even the best gravity - 
filters are not perfect and yet they are in every re- 
spect better for a city than the individual use of any 
ordinary domestic filter; for most every pattern in 
the market is worse than none and vastly inferior to 
the employment of boiling for purifying water at the 
time of an epidemic. 

Skilfully constructed sand-beds, usually an acre 
more or less in area, are capable of purifying river 
and other surface waters very acceptably, provided 



28 

there has been formed upon the surface of the fine, 
top sand layer a delicate and continuous film of con- 
fervoid vegetation, silt, and other fine particles which 
occur in such natural waters as require filtration. 
Since the first half of this century practical English 
working has demonstrated this very completely. This 
principle was introduced into Germany during 1856, 
but the true merit of the process and the essentials 
upon which its great hygienic value depends were not 
fully appreciated till more than a qurter of a century 
later. It was carried out in a very successful form 
at the Tegel river- lake supply of the Berlin water- 
works in November, 1883, to remove the trouble- 
some, pipe -obstructing growths of crenothrix which 
resulted from the admission of deep well-waters. 
There the process has been studied most carefully 
since that time, and the thorough chemic and bio- 
logic results, noted by Proskauer and others, have fur- 
nished an invaluable guide to the understanding of 
the conditions which underlie success with such fil- 
ters. So, too, have the corresponding investigations 
by the English Local Government Board and by the 
Massachusetts Board of Health, whose results are 
given in detail in their reports published during the 
last seven years. 

It is found that, unlike sewage filters, the sand- 
beds employed to purify water for public supply do 
their important work chiefly by virtue of the upper 
layer of fine sand, and particularly by reason of the 
presence of an unbroken, fine film allowed to settle 
upon the surface of the sand before the filtering be- 
gins. In proportion to the perfect quality of this 
very fine straining medium is the efficiency of the 



29 

filter as regards the separating of all (or practically 
all) bacteria from the water that passes slowly 
through the filter-bed. This removal of bacteria is 
the criterion of hygienic merit. As the flow through 
is at the slow rate of not much more than four inches 
an hour, considerable chemic purification of the 
water takes place during its downward passage 
through the masses of sand and gravel. On the im- 
mense area of sand and gravel surfaces, bacterial 
slime (zooglea), containing great multitudes of nitri- 
fying bacteria, is attached. As the water, contain- 
ing one per cent, of oxygen or more (if quite cold), 
flows slowly over these, the putrescible organic mat- 
ters present are, to a varying extent, still further 
converted into inorganic compounds. This chemic 
change is due to the vital activity of the permanently 
adherent nitrifying bacteria whose presence, especi- 
ally at the top, makes an old filter more efficient in 
this respect than a new one; yet, as Gill and 
Pief ke first showed, simple washing does not cause 
sand to lose this zooglea slime. Before the para- 
mount importance of the separation of bacteria was 
recognized, greater stress than now was laid upon 
determining the degree of this chemic action. At 
the most, the chemic improvement is, as Frankland 
expresses it, ' ' slight, ' ' and is not of much hygienic 
importance as compared with the bacterial efficiency 
of the film on th^ surface. 

Suppose, then, a plant carefully constructed. It 
has a settling-basin large enough to hold several 
days' supply of crude water. Each filtering-basin is 
an acre or more in area; the firm walls and especially 
the bottom are water-tight by means of several 



30 

inches of good concrete. There should be no venti- 
lating pipes rising through the bed, and no projec- 
tion or irregularity on the walls to restrict the facility 
of maintaining a continuous, smooth surface of sand. 
That of the upper portion of the bed is very fine, 
although it is not found best to have the grains less 
than one- fourth millimeter (j^ of an inch) in di- 
ameter. Two feet or much more of fine, washed 
sand (although some have even less than a foot and 
yet work very satisfactorily, as at Alexandria, Egypt*), 
are supported by coarse sand, and this by gravel 
and broken stone. This makes the filter about six 
feet deep. In some the sand rests on two courses 
of porous brick, and is reported to work well. The 
writer has not tested such thin filters. After filter- 
ing for from one to eight weeks, according to the 
condition of the crude water supplied, the sand sur- 
face will have collected so much slime and dirt that 
it is clogged, and water passes through only very 
slowly under any safe pressure. Then the inflow is 
stopped and water is drawn off through the outflow 
sufficiently to render the sand firm. Practised em- 
ployees, having broad boards attached to the soles 
of their boots, go over the surface systematically, 
shoveling off the upper half inch or more of the 
fine soiled surface -sand together with the accumu- 
lated sediment above it. This is then removed to a 
neighboring washing arrangement. If this scraping 
has caused any disturbance of the even surface of the 
sand, this must be completely rectified, then all 
made smooth and perfectly level. At times more 
sand has to be added very carefully to replace that 

» Ann. d. Hyg, Publ,^ 1895, p. 487. 



31 

taken away by repeated scrapings. It is largely to 
obviate this that the sand- layer is made several feet 
thick at first, for the bacteria-separating efficiency 
may be equally good or nearly so if even less than a 
foot of fine sand be present, yet the brief increase in 
the number of bacteria, with the effluent which comes 
just after cleansing of the surface, is greater in shal- 
low than in deep filters, probably because of me- 
chanical disturbance of the bed in its deeper parts. 
The majority of experts think five-foot layers of sand 
yield better results than very thin beds. 

After being cleaned and leveled, the filtgr-beds 
should have some of the purified water returned 
through the outflow pipe until the surface of the 
sand is just covered. Then crude water is gently 
introduced, through the regular supply-pipe, until 
the ordinary working height of three or four feet is 
reached. Care taken to accomplish this refilling 
without any disturbance of the sand surface is repaid 
by superior efficiency of the filter from the commence- 
ment of working, and by these means little or none of 
the first effluent may have to be wasted on account 
of containing too many bacteria. Before any water 
flows through, that carefully introduced should be al- 
lowed to stand for twenty -four hours. At the end of 
that time the all-important film will usually have 
formed upon the sand surface. If the sediment be 
insufficient, some coagulant, such as alum or iron salts, 
may be added to produce this; yet the natural film, 
especially if of minute earthy particles, works best. 
When once the film is sufficient to separate bacteria 
properly, as determined by the usual gelatin -culture 
test of samples allowed to flow through, but not yet 



3« 

turned into the hydrant supply, the filter may begin 
to work. The impure water above the filter-bed is 
constantly maintained at the same level, not exceed- 
ing fifty inches, and the outflow is automatically 
regulated so that not much more than four inches 
flow through the sand each hour. That gives a yield 
of not more than from two to three million gallons 
per acre in twenty-four hours. This is the highest 
safe yield generally considered practicable. Yet in 
the Twenty-seventh Report of the Massachusetts 
State Board of Health (1895), pages 517 and 518, it 
is stated that upon experimental filters of very small 
size (^f-jy-J^nr ^^ ^^ acre) filtration at the rate of from 
five to nine million gallons daily has been shown 
capable of removing more than ninety-nine per cent, 
of all bacteria! The unfiltered water contained 
several thousand bacteria, being from the Merrimac 
River, which at I>awrence contained considerably 
less than 30,000 bacteria per cubic centimeter, the 
average being over 16,000. The sand was five feet 
deep and 0.38 millimeter (y^l^ of an inch) was its 
** effective size," this term indicating the ** max- 
imum diameter in millimeters of the finer ten per 
cent, of the sand grains." (See Twenty-fourth and 
Twenty-sixth Reports, pp. 541 and 703, respectively.) 
Intermittency in operation, while necessary for 
sewage filtration, is both practically and theoretically 
less satisfactory than constant operation without 
cessation from one cleaning -time to another. This 
is especially so during winter. These filter-beds 
work continuously, then, when once the water is 
started running through them. The clearer the crude 
water, the longer the interval between cleanings. 



L 



33 

It is an advantage to employ settling -basins as large 
as practicable as a preliminary to the filtering proc- 
ess. They are indispensable when the supply is very 
turbid, such as are, notably, the waters of the Ohio 
and Missouri rivers. Gravity sand filter-beds, prop- 
erly conducted and carefully managed, afford, on 
the whole, the most reliable artificial means of ren- 
dering an impure water fit for both domestic and 
general consumption. At their best they not only 
clarify the water, but also reduce the average num- 
ber of bacteria to considerably less than one per 
cent, of the amount in the original crude water. At 
the same time, practically all disease bacteria are re- 
moved, yet some filters yield very poor results, whether 
because of incompetent management or through in- 
herent defects in the construction, or by forcing a 
sand-bed to produce beyond its capacity. When 
the surface is frozen the results are more apt to be 
defective. 

The cost of constructing a complete filter-bed an 
acre in size varies from $20,000 to $100,000, ac- 
cording to the topography, the availability of suit- 
able sand and other material, the cost of labor, and 
other elements. An open filter-bed, having tight 
concrete bottom and sides, with proper regulators 
and piping, not including settling or receiving 
basins, may roughly be said to cost at least $1 for 
every square foot of filtering surface. Half as much 
again must be added for a suitable covering to pre- 
vent the growth of algae and keep the water of more 
even temperature. In northern localities some cov- 
ering is desirable because of the considerable im- 
pairment in efficiency caused by ice during winter. 



34 

At Hamburg, open filters are employed and success- 
ful cleaning of an ice-covered filter is accomplished, 
without the usual removal of the water, by using a 
sort of trawl scraper suspended at the right level by 
floats and moved across the sand-bed. This is ex- 
plained and very well illustrated in the Journal fur 
Gasbeleuchtung and Wasserversorgungy 1897, page 
4, a portion of the illustrations and summary of 
which appeared in the Engineering Record for May 
I, 1897. 

When suitable fine sand is not procurable in the 
neighborhood, the expense of procuring it may be 
too great to warrant its use. Crushed quartz, espe- 
cially adapted for sand-filters, costs at the rate of 
more than 60 cents per cubic foot, weighing about 
ninety pounds. When the cost of the sand is, thus 
prohibitive, and when ground enough properly lo- 
cated is not to be obtained at a reasonable price, 
recourse may be had to mechanical filters, which 
occupy much less space and require much less sand, 
since water flows through them fifty or more times 
as rapidly as through the same area of gravity sand- 
bed. While these are more expensive to construct, 
the operating expense is much less than that of 
mechanical filters, being mainly that of cleaning 
and restoring the surface. It may be roughly esti- 
mated at nearly ;j;io per million gallons of filtered 
water. It is claimed that a few plants require much 
less than half this outlay, while others cost half as 
much again. At Lawrence, Mass., the cost was less 
than %i per million gallons during 1895. Nearly 
forty per cent, of the expense was for removing ice, 
while less than half was for scraping and replacing 



35 

sand. Nearly fourteen per cent, of the charge was 
for washing the removed sand. 

Filter-galleries, situated in the low banks of rivers, 
are fed rather by ground waters from the upland side 
than by filtration of river water, which does not filter 
with practicable speed * through the very resistant 
mud and silt. When the river rises and its water 
flows into the galleries through coarse sand and 
gravel, it is likely to render the supply .less whole- 
some by reason of the ease with which bacteria can 
pass through. Thus, at Dresden, where a ground 
water-supply of this sort is used for a demand a 
third larger than originally planned for, it is found 
that when the Elbe rises more than a meter above 
its normal level, the bacteria of the water (usually 
twenty to forty per cubic centimeter) increase to 
several thousand. With this increase a greater 
number of gastro -intestinal disorders than usual 
occur, especially among infants, and their mortality 
is increased (Meinert). For such filter-galleries it 
is essential to have water-tight barriers constructed 
to keep out any inundation or any inflow of water 
through ground that is not of a perfectly filtering 
character. 

While figures may prove very misleading when not 
carefully interpreted in the light of the exact condi- 
tions obtaining, it is not uninstructive to give a few 
statistics to illustrate the hygienic value of good fil- 
tration. The great cholera mortality (8200) at 
Hamburg during 1892, while the similarly situated 
neighboring city, Altona, using worse water from the 
same much-polluted river was comparatively free, 
was because the latter had good filtration, while at 



36 

Hamburg the old filters were defective and the new 
ones not quite finished. The typhoid mortality there 
is only 58 deaths per 1,000,000, while at Altona it is 
41, and at Berlin, between 40 and 65. Amsterdam, 
filtering water from sand lakes or canals, has but 
28 deaths per 1,000,000. 'These figures compare 
favorably with those of cities supplied with first-class 
mountain water, Munich having 32, Vienna 40, and 
Nuremberg 42 deaths from typhoid per 1,000,000. 
As against the results of these naturally or artificially 
pure waters, may be quoted Maline, supplied by 
shallow surface wells, the mortality there being 993.* 
This, however, is exceeded by the Chicago typhoid 
death-rate for 1891, which Hill gives as 1600. He 
gives that of Pittsburg (for 1893) as mo, St. Louis 
(1893) as 1030, and of Lawrence, Mass. (1892), 
1020. 

Artificial sand-plate filters (Fisher system) are 
said to be employed with great success at Worms, 
Westerhiisen (near Magdeburg), and at the Muggel- 
lake intake near Berlin. The writer has never tested 
these, but credible reports would make them seem 
about equal to good gravity filters as regards bacteria- 
removing efficiency, while the yield is proportionately 
much greater. If thorough tests, carried out im- 
partially at the last -mentioned station, confirm the 
excellent results claimed elsewhere, it may well be 
believed that this method, which seems the revival 
of an older idea, is practicable on a large or small 
scale. The plates are easily and quickly cleaned 
and, like mechanical filters, have the advantage over 
gravity sand filter beds that they can be sterilized 

> Smee, Lancet^ March 13, 1897. 



37 

promptly and completely by steam in case of typhoid, 
cholera, or other infection having entered They 
are very strong and have the especial merit of sim- 
plicity and freedom from increase of bacteria at the 
beginning of the flow just after cleansing, as also un- 
der a considerable increase of pressure. This would 
mean that attempts to overwork them, so common 
with mechanical sand-filters, would not produce such 
bacteriologically bad water as results from the latter 
under such circumstances. They are said to be com- 
posed of fine sand and finely ground soda-water- 
bottle glass mixed and heated to 1200° C. The 
slabs are nearly forty inches square and four inches 
thick. They are screwed together and the edges 
cemented. A number of them may be arranged in 
line to form a pipe into which water filters from 
without. 

In some cases in which filtration is deemed neces- 
sary for a municipality, an industrial plant or even 
individual systems, the slow, gravity -sand-bed system 
above considered appears impracticable, or less de- 
sirable than mechanical filtration, even for a large 
supply. This may be due to remoteness from a 
supply of good sand, or the ground may be limited, 
or other economic reasons may obtain. Whatever 
make be selected, even including the invertible pat- 
terns as that of Krohnke or of Carl Schultz, it will 
be quite similar to its numerous rivals. These me- 
chanical filters operate by causing the crude water to 
flow rapidly, usually under hydrant pressure, through 
comparatively small masses of fine sand, coke-pow- 
der, or similar substance enclosed in cylinders of 
wood or iron. Mere sand (or coke, etc.) alone will 



38 

not be capable of holding back many bacteria, nor 
will it strain turbid water sufficiently clear, but the 
preliminary addition of one or two (or more) parts 
of a coagulant, such as an alum or iron salt, to one 
hundred thousand parts of water usually causes the 
formation of a gelatinous film which, settling within 
an hour upon the surface of the sand and also to a 
certain extent between the grains, when skilfully 
managed will cause such a filter to produce clear 
water which compares favorably with that from good 
gravity sand-filters. A permeable bottom supporting 
the sand allows water to flow downward into the out- 
let and also permits purified water to be driven 
backward for the (from one to five) daily washings. 
Upon this a layer of coarse sand or crushed quartz 
about a foot deep should be placed. This supports a 
two-foot layer of find sea sand or crushed quartz (or 
coke, etc.) of an effective size of from .3 to .5 mm. 
(from -g^ to ^ of an inch). Into the space above this, 
crude water for purification is introduced after alum 
has been added. 

Both as an expert trying to aid in obtaining the 
best possible results, and also as an impartial scien- 
tific observer wishing to determine how widely such 
filters vary in efficiency, the writer has tested many 
of these. While his investigations have been less 
complete than those conducted at Providence, R. I., 
during 1893, and at Louisville, Ky., during 1896, 
as regards thorough and constant study with one kind 
of water, they yet serve to show the results accom- 
plished by such filters under ordinary as well as expert 
management in actual every-day use under practical 
working conditions. Tested regularly and irregularly 



39 

in this way, mechanical filters generally show less 
constant and perfect results than do gravity sand 
filter-beds, but this is more owing to lack of carefiil 
management and insight into the principles of oper- 
ation than because of any inherent defect in the 
method. Almost all of the various makes will do 
excellent work, clearing muddy water and reducing 
the number of bacteria to less than one per cent, of 
the original amount; and yet some of the best makes, 
as typified by filters established in permanent opera- 
tion, and referred to in the makers' circulars, have 
been found to produce an effluent having more than 
fifty per cent, of the bacteria (of all varieties) occur- 
ring in the crude water. On the other hand, some 
of them at times are found, without any unusual care, 
to produce water almost entirely free from bacteria, 
provided they are worked at much less than their full 
capacity. Others, when overworked, offer insufficient 
resistance to bacteria and turbidity. They are com- 
paratively complex and usually dependent upon con- 
stant skilled supervision. Their all-important bac- 
teria-separating film, unlike that of the surface of 
gravity sand-beds, not lasting for weeks perhaps when 
once established, but rather being destroyed by the 
necessary washing at least once daily, and perhaps five 
times, the operation cannot be so well controlled as 
that of the slow filters. With the latter, the yield 
may be stored in a separate basin until the usual 
gelatine test shows whether the film is separating bac- 
teria properly. But with mechanical filters the es- 
sential filtering-film must have been removed, as also 
a number of subsequent ones, under conditions of 
economic working, long before a bacteriologic test 



40 

can have been completed. This is a very important 
consideration especially at times of epidemic and 
actual infection of the water. 

A point in favor of mechanical filters is that they 
can be completely disinfected by steam or boiling 
water, the most satisfactory of all means. Within 
these filters the sand does not remain in as perma- 
nently good working condition as in those parts 
of gravity sand-beds which lie beneath the sur- 
face. Accordingly, it needs to be steamed or else 
scoured with a four- or five-per-cent. solution of 
caustic soda, or both processes may be employed, at 
least twice yearly. Otherwise, bacteria, not removed 
by the daily washings, grow in time so numerous as 
to contaminate the effluent, and among these may be 
disease germs. If their presence be suspected, and 
certainly if demonstrated by culture-tests of the sand 
or water, sterilization of the filter is necessary. Me- 
chanical filters involve considerably greater expense 
for operating than do gravity-beds. On the other 
hand, the plant costs less than the best water-tight 
slow sand-beds complete. Although alum costs less 
than two cents a pound, this proves a costly item 
when used for large quantities of water. When the 
filter is working well, the proportion of alum used 
may be only one part to a hundred thousand parts of 
water, or less than half a grain to the gallon. In 
practice, however, this proportion of the coagulant 
is often considerably exceeded, whether through 
carelessness or because of very muddy or peaty qual- 
ity of the water, and this of course increases the ex- 
pense. From such permissible small amounts of 
alum as remain unchanged in the purified water we 



41 

cannot recognize any danger to health or detrimen- 
tal influence upon boilers or industrial processes. 
This alum is almost entirely decomposed in making 
the desired film of flocculent alum hydrate and of a 
combination with organic matter present. The sul- 
phuric acid set free combines with the earthy carbon- 
ates present and thus tends to become neutralized. 
At the beginning of filtration, after the film and dirt 
have been washed out, an excess over the usual mi- 
nute amount of the coagulant is commonly added to 
hasten the formation of the necessary film. What- 
ever of the salt remains uncombined is. washed out 
with the first few minutes' flow of water. Since this 
water has very many bacteria present, it is purposely 
wasted from well -managed filters. This amount of 
water wasted varies from three to eight per cent. 

While only vertical cylinders are spoken of above, 
it should be mentioned that horizontal cylinders are 
in some cases employed in large plants, yet both are 
the same in principle. Even an invertible one, 
such as that of Mr. Carl Schultz, or like those used 
for the KrQhnke deferrating process, work in prac- 
tically the same way, the peculiar construction being 
only a modification designed to facilitate cleansing. 
Mechanical filters may be combined to any number 
to make *' batteries " and thus millions of gallons be 
produced by one plant. Individual filters vary in 
size from one foot to ten feet in diameter. Since 
the water is meant to flow through nearly sixty times 
as rapidly as through a slow gravity filter-bed, a fil- 
ter ten feet in diameter, and having an effective fil- 
tering surface of about seventy-eight square feet, 
would be expected to yield from loo to 250 gallons 



42 

per minute, but should not be allowed to exceed 
that amount. Filters one foot in diameter are cap- 
able of yielding loo gallons per hour, and in com- 
plete working order cost not far from $ioo. They 
are intended for private houses and serve the pur- 
pose of clearing all the water for the household under 
ordinary conditions, but they cannot respond to sud- 
den demands for a large quantity of good filtered 
water as the combined draft for laundry, and other 
purpose, coming at the same moment would produce 
unsatisfactory water. Accordingly, storage-tanks are 
needed, and wood will usually be selected as the 
material. This, and also the filter, usually do not 
receive careful attention and the water is then too 
often likely to yield many bacteria when tested. 
Pure drinking-water would be better insured by a 
battery of Berkefeld filter-tubes kept in good order 
and sterile, which, even at present high prices, would 
prove less costly. A drawback to some of the best 
varieties of mechanical filters is threatened litigation 
if the coagulant be added just as the water enters the 
filter. The Hyatt patent, controlling this, expires 
within four years. The use of a preliminary settling- 
tank or basin is not thus restricted, but seems less 
satisfactory. 

Sedimentation is of recognized value in reducing 
the number of bacteria as well as tubiditv of a mass 
of water, which can remain at rest or nearly so as in 
storage-basins, and in lakes, or lake-like enlarge- 
ments of rivers, when the current is very sluggish. 
Various bacteriologists have found this to be the case, 
and the writer's studies of lakes and reservoirs, espe- 
cially when much silt was present, have tended to 



43 

confirm this view. The presence of settling-basins, 
with or without aerifying arrangements for the in- 
flow, is of value as an auxiliary, but is not of enough 
absolute potency to warrant exclusive reliance upon 
such means. In most of our streams and bodies of 
water which receive bacterial contamination from 
sewage and other filth, sedimentation does not seem 
to accomplish much. It is, as already said, the ele- 
ment of great dilution which seems to be the chief 
factor in improvement of the water. Light (which 
favors the production of algae) lessens bacteria, espe- 
cially when the water is well aerified, according to 
most experiments conducted in that directions; and 
strife among bacteria is regarded as aidmg the pro- 
cess. If a small volume of infectious washings from 
a privy- vault, manured field, or sewage outlet, mingle 
with abundant water of a large stream, the tendency 
to diffusion throughout the large volume of water 
causes a sample taken lower down stream to show 
many less bacteria, as well as greatly diminished 
chemic impurity, unless some intervening contami- 
nation complicates the conditions. In small streams, 
flowing through densely peopled industrial districts, 
this natural purification is so much less potent that 
English investigators have declared that a stream 
much polluted cannot again become pure. In our copi- 
ous rivers, as also on the Continent of Europe, such 
an assertion is easily disproved, yet the experiences 
of some of our cities show that extreme vigilance 
must be exercised if the water-supply comes from a 
contaminated river. An arbitrary statement that 
fifteen or even fifty miles of flow insures the destruc- 
tion of all harmful contamination is not always sub- 



44 

stantiated by facts. In this connection, although 
the same principle has not been shown to be opera- 
tive in our streams, it seems proper to mention Han- 
kin's observations on the peculiar quality of the water 
of the great rivers of India*, in which certain acid sub- 
stances, volatile on heating, destroy all cholera ba- 
cilli that enter the streams, although these danger- 
ous germs increase vigorously in well waters of those 
regions. 

Observation shows that storage for months tends 
to lessens or remove color and odor from water; but 
when light is not excluded, algae are likely to flour- 
ish, if nitrogenous elements in an assimilable form, 
and special mineral constituents, and other condi- 
tions, are present to favor their growth. Even when 
such storage is practicable, it is always preferable to 
have water in the beginning so pure that none of the 
causes of odors, such as infusoria, sponges, nostoc, 
etc., or any dead animal or vegetable matter, are 
likely to occur. Although such waters are not known 
to have directly caused disease, a fishy, or other re- 
pellant odor or taste, can make a water so offensive 
that people will not drink it, and it will arouse more 
emphatic protest than even the known presence of 
disease -causing germs. If a reservoir be created by 
damming up and overflowing fertile slopes or mead- 
ows, the soil should be dug away so as to remove 
roots of grass and bushes, may be to a depth of more 
than two feet; otherwise, decomposing organic mat- 
ter may cause bad odors to arise in water stored 
there. 

Heat kills the germs of disease and all other living 

^New York Med. Jour. ^ April lo, 1897, p. 503. 



45 

things. If water be boiled long enough, no bacteria 
or other organisms will remain alive in it. This, in its 
practical bearings, has long been recognized. The 
writer pursued a prolonged and careful series of ex- 
periments some years ago in this country (chiefly in 
the Carnegie Laboratory) and in Koch's laboratory 
in Berlin, employing various natural and artificially 
infected waters, and demonstrated that simply bring- 
ing water to nearly the boiling-point, and then allow- 
ing it cool gradually, sufficed to destroy the elements 
of infection. Thermophilic bacteria and some com- 
mon kinds require very much longer heating for their 
annihilation, but practically a brief heating answers 
the hygienic purpose. This means can be employed 
almost anywhere and is to be recommended at least 
as a makeshift under urgent circumstances, as when 
the water-supply is bad or even suspicious and yet 
no adequate purification obtains. Since boiling 
drives off air and also carbonic-acid gas (causing the 
separation of lime and magnesia), the water in most 
cases is thereby made somewhat less palatable; al- 
though the taste of some malodorous and unpalatable 
waters is improved by heating. Subsequent aerifica- 
tion, as by use of a bellows or more elaborate appli- 
ance for exposing the water to as pure air as possible, 
effects an improvement. The addition of a small 
amount of tea improves the taste for some people. 

Within the last ten years several economic con- 
trivances for heating water to the boiling-point and 
cooling it to within a few degrees of its original 
temperature have been perfected by Yaryan, Siemens, 
Grove, Rouart, and others. The apparatus of the 
first-named inventor, as designed for the use of a 



46 

large mimic L|iality, was carefully tested by the writer 
in Toledo and found capable of completely sterilizing 
the hydrant water of that city. The principle em- 
ployed Ijy all these independent inventors is a series 
of pijies so arranged that by conV^ction the cold in- 
flowing water cools the outflowing sterilized stream 
and is at the same time gradually heated before it 
arrives at the boiler. Thus a minimum expenditure of 



fuel is required. The accompanying illustration (Fig. 
4), supplied through the courtesy of Rouart Freres of 
Paris, shows an apparatus which is claimed to have 
achieved a practical success. The arrows indicate 
the course of the water from the inflow E, to the out- 
flow F. The yield is of better quality than non- 
aerified, boiled water, since not much more than 
half of the oxygen is lost in a well -constructed ap- 



47 

paratus. Lime and magnesia are considerably re- 
duced, as also are the other mineral salts and the 
organic matter. Despite the theoretic merits of 
these economic sterilizers for water which is clear 
but infected, they have not found practical accept- 
ance unless in a very limited way. 

Distilled water is considered desirable for the most 
economic and satisfactory results in certain processes, 
as also for human consumption, when the natural 
supply is of poor quality, and especially when it is 
charged with poisonous or other mineral salts. To 
supply this need at a low cost, on a large scale, ex- 
cellent stills exist. On shipboard, particularly when 
the vessel is on a prolonged voyage or cruise away 
from ports where good, fresh water can be obtained, 
distilled water, easily prepared and palatable, is not 
only very convenient as saving the space of storage- 
tanks (too often defective) and casks likely to become 
foul, but is also of considerable sanitary value. This 
is evidenced by statistics of the naval and marine ser- 
vice before and after the introduction of the regular 
use of the Baird, Niemeyer, and other excellent stills. 
All properly constructed ones yield water free from 
bacteria if due care be taken to have the condensing 
and outflow portion and any filter there sterilized by 
steam before using. Instructions should also be given 
to waste for ten or twenty minutes the first flow of 
water after long disuse of the apparatus. This is to 
avoid chemic impurities. Lead should never enter 
into any part of the construction of tanks or stills, 
for the danger of poisoning is considerable, even 
from solder. Tin shown by assay to be pure should 
be used for joints of tin plates and tubes, although 



48 

more difficult to work with than solder. Iron rusts, 
and is, therefore, inferior to well-tinned copper plates 
or drawn brass coils well tinned inside and outside. 
Abundant sea-water cools the steam coils by flowing 
in at the bottom and out at the top on one side of 
the tight casing inside of which are the smooth or 
corrugated coils of vertical pipes into which steam 
under low pressure enters at the top. At the lower 
end of the condenser the distilled water flows out or 
is drawn out by a pump where a filter (of asbestos or 
pure bone, charcoal, and sand) is present, for the 
purpose of relieving the water of the whitish flecks 
and oily taste which come from slight quantities of 
lubricating oil going over with the steam. When 
the steam pressure is high this oily taste is more 
likely to occur than when the pressure is quite low, 
as from an evaporator especially introduced for the 
sole purpose of vaporizing sea or other water in order 
that it may, on condensation, become distilled water. 
From a perfect still very pure distilled water can be 
produced, having none of the salt or other undesir- 
able constituents of the original water. If petroleum 
lubricants are used, they tend to cause an increase 
of ammonia above the small amounts which can be 
detected in almost all distilled water. At the end 
of every voyage the filter attached to the condenser 
of a vessel's still should be cleansed and sterilized by 
steam sent through the still. The distillate should 
also be tested from time to time to detect any leak- 
age of the cooling salt water into it. One need only 
add a drop or two of a solution of silver nitrate, 
slightly acidulated with nitric acid. If any salt has 
entered the distilled water, this reagent will produce 



49 

a visible whitish cloud or precipitate when added to 
the water in a glass. 

Aerification of distilled water should be so managed 
as to introduce no bacteria, and certainly no harmful 
ones. Previous filtering of the raw water and boil- 
ing long enough to insure the driving off of the more 
volatile organic matter previous to distillation is re- 
garded by Mr. Hallock, the expert chemist of the 
Schultz establishment, as the best means of prevent- 
ing the usual taste of ordinary distilled water, and 
more satisfactory than the preliminary addition of 
such an oxidizing agent as permanganate of potash. 
For the Witteman still the water is twice boiled and 
superheated before being vaporized. Thus, for arti- 
ficial ice or artificial mineral waters the maker who is 
careful enough to use distilled water avoids the neces- 
sity of aerifying it, which process too commonly al- 
lows the introduction of bacteria of numerous kinds 
which may be in the air, and which, despite cotton 
and other filters of the air, are always found in waters 
of most establishments. 

There are localities, notably in desert regions, 
where the available supply of good drinkable water 
is so deficient that the providing of distilled water in 
large or in small amounts is a hygienic necessity. 
Aden (steamer point), on the Arabian coast, affords 
an instance where large evaporators of American and 
other types are employed for supplying water. Es- 
pecially when the apparatus uses an exhaust pump is 
it important to test for the presence of lead contami- 
nation or its probability. In our western mining 
regions and in certain grazing districts in the ''Great 
American Desert," water is often so impure and un- 






so 

palatable as to require distillation in order to be ren- 
dered fit to drink. When the supply is copious 
enough to permit an abundant flow of cold water for 
keeping condensing-coils or plates cool, use may be 
made of the common form of glass or tin spiral tube 
connecting tightly with a large closed kettle or ves- 
sel in which water is kept boiling. The resulting 
steam goes out of a small opening into a long pipe, 
which rises at first and then, like the coil into which 
it conducts the steam for condensation, descends 
gradually. The spiral tube or coil is fixed in a pail 
of freely wasted cooling water, and conducts its con- 
densed steam as distilled water outside by passing 
through an opening on the side of the pail. By 
bending this distilled-water tube upward and then 
downward so as to form a siphon after it emerges, 
contamination by leakage of the cooling waste water 
is obviated. Such a primitive arrangement produces 
distilled water that is sterile if only the inside of the 
condensing-pipe be sterile beforehand. So too do 
tinned copper drums or pipes through which water is 
condensed from steam in an automatic way, the air 
outside these large -surfaced conductors cooling them 
sufficiently to secure condensation. Attention is 
needed only to keep boiling water in the kettle. The 
joints should be quite tight. 

During the last ten years the writer has seen vari- 
ous adaptations of this principle of using cool air for 
condensation, and those which he has tested have 
proved capable of yielding wholesome water, free from 
bacteria and mineral matters, except that lead has 
been evident when even a little solder has been used 
for the joints. This is especially the case when the 



5' 

apparatus is new. If such a faulty new condenser 
must be used, the first few gallons ought to be 
wasted, and also some after disuse. Tests show that 
it is safer to tolerate a small leakage of steam 

FlO. s. 



Dumeslk- rondensor. Arrow indicales direction of inflow of 
steam. DisllLled water flows out at base. 

from a well-made condenser than to use solder, in- 
stead of tin and the blow-pipe, for stopping it. The ac- 
companying illustration (Fig. 5 } shows the appearance 
of a domestic condenser of the most effective kind. 



52 

These are usually made as a series of drums commu- 
nicating by short pipes with tight joints, as illustrated 
by Sears in the Nav York Medical Record, March 
2 1,1 896. Practically the same idea is embodied in 
a subsequent model used for the ''400" still, which 
has a cooling surface of over ij4 square meters. 
These stills yield from a quart to a gallon per hour, 
according to their size, and somewhat according to 
the outside tempeiature and the way they are worked. 
The steam kettle may be heated on a stove, but a 
more constant heat can be obtained from a central- 
draught ** mammoth" lamp or an ** oil -gas" stove 
or a gas stove. When gas is supplied under suffi- 
cient pressure, and when water abounds for cooling 
the steam, a still of the Jewell pattern or the '* Do- 
mestic" style can produce a gallon or more per 
hour. The writer has used the latter kind during 
the last nine years, and found it satisfactory. There 
are various American and foreign domestic stills 
employing a comparatively small tinned surface for 
condensing, and flowing water is used (and also ice) 
for cooling this. They are generally less satisfac- 
tory than the styles above indicated, although they 
produce distilled water of fair quality. 

Chemicals, in which class may be included alcohol 
in beverages and also the employment of electricity in 
any form or adaptation, are not to be recommended or 
relied upon for ordinary practical sterilization of drink- 
ing-water. It is true that various reagents of this class 
if acting in sufficient strength and for a sufficient length 
of time are capable of destroying bacteria in water. 
We know that, by admixture of just the proper 
amount of certain chemicals, foul water and even 



53 

sewage can be clarified and even sterilized. The 
importance of the skilful use of quicklime under cer- 
tain conditions for the removal of organic matter or 
for lessening the amount of lime already present, is 
recognized. Like alum, iron, or similarly acting 
coagulants, it has a distinct utility when indicated for 
the purification of bad waters, but the use of these 
chemicals, or of permanganate of potash or such chem- 
icals asbromin (urged by Schumburg), or any others 
in sufficient quantity to produce immediate steriliza- 
tion of bad water in all cases, may make it anything 
but wholesome or palatable. Boiling is in every 
way more reliable and therefore preferable. 

It is quite possible with time to produce, under 
laboratory conditions, the death of all the bacteria 
of a given sample of water, or at any rate cholera 
bacilli, if even so mild a natural acid as lemon juice 
is added to it. Stronger agents are of course more 
potent, but it is unwise to recommend that such un- 
certain means be relied upon by the masses in time 
of epidemic. Some abominably tasting and rather 
harmful drugs are at times thus advised without giv- 
ing the authority and without retraction. Even in 
serious technical literature fallacious statements of 
assumed fact are encountered. Thus, Traube as- 
• serted that a minute amount of lime chlorid sterilized 
water. Although obviously erroneous, the statement 
attracted some attention because of the high charac- 
ter of the organ of publication. Various tests showed 
that at least thirty times as much as the amount 
specified was needed. To sterilize bad water in ten 
minutes one must use of lime chlorid no less than 15 
parts to 100,000 parts of the water. Even more 



54 

may be refjuired if the water contains much organic 
matter, for that takes up the chlorin so rapidly as 
not to leave enough to act on the bacteria. The 
lime chlorid is not easy to dissolve and a sufficient 
quantity may make the water very hard as well as 
cloudy, thus necessitating filtration. So, too, the 
alum process urged by Babes was proved to be in- 
adequate. Although in some cases 4 parts of alum 
to 10,000 parts of water prevent increase of bacteria, 
the writer determined from repeated experiments 
that under practical conditions even this consider- 
able amount could not always be relied upon. Per- 
oxid of hydrogen acting, like ozone, because of its 
contained active oxygen, if in the proportion of i 
part to 1000 of water, sterilizes it in twenty-four 
hours, but when organic matter is present this agent 
is not to be relied upon. All such agents act in too 
complex and uncertain a way to be comparable with 
boiling. 

Electricity deserves mention chiefly because of the 
prominence which has been given to it. If con- 
ducted through water, a dissociation of the molecules 
into the component hydrogen and oxygen occurs, 
the latter being set free at the positive pole, and it 
can exert a destructive influence upon bacteria, es- 
pecially in sunlight or strong daylight. This affords 
an interesting laboratory demonstration, but it costs 
too much for ordinary employment and is otherwise 
impracticable. Electrolysis of common salt, present 
in water, causes the salt to break up into sodium and 
chlorin. These tend to recombine and one of the 
products is chlorinated soda, just as chlorinated lime 
results from electrolysis of lime chlorid. By reason 



55 

of their available chlorin, these resultant salts are 
powerful disinfectants and are known to be destruc- 
tive of bacteria if enough of the fresh, powerful 
chemical be present, but this is merely a chemic ac- 
tion and no, other and subtler force than the chemic 
products ate- operative in the process. Because of 
obvious uncertainty as to the exact quantity of these 
evolved chemicals when they are introduced in this 
indirect way, it is preferable that they be added in 
a precisely known quantity. This is also the more 
economical way, although the use of such chemicals 
cannot be regarded as a desirable means of sterilizing 
drinking-water. 

It should be mentioned that when lime is used in 
the largest permissible amounts to clarify sewage, 
the effluent water is almost never free from bacteria, 
and in no case is wholesome even if used at once. 
If stored or poured into a river, its alkalinity favors 
the growth of algae and other forms of water vegeta- 
tion. The effluent from a properly managed sewage- 
farm is much more suited for admixture with water 
of streams that further on are used for water-supply. 
Natural filtration by the soil or efficient artificial fil- 
tration must be considered — like boiling — as vastly 
superior to chemic and electric means of sterilizing 
water. Any insoluble substance in minute particles, 
which is clean and just heavy enough to sink slowly 
in bacteria-laden water through which its innumer- 
able grains or flakes are diffused, will in settling 
carry down a considerable percentage of the bacteria. 
Even rapidly sinking particles, such as grains of 
sand, can experimentally be shown to produce this 
result to a certain extent although less completely 



S6 

than particles which settle more slowly. The gela- 
tinous sediment resulting from the addition of alum 
or other coagulant is very effective in this respect. 
So, too, on adding fresh lime and also soda to re- 
duce the hardness, the fine resultant precipitate in 
settling causes many bacteria to sink with it. But 
this removes only a portion of the bacteria, for it is a 
mechanical removal and not a chemic action, unlike 
the total destruction caused by adding an excessive 
amount of fresh lime or other strong chemic agent. 

Hardness of water varies according to the amount 
of lime and magnesia salts present. Distillation of 
water, heavily charged with these or other salts, 
yield a product free from them, and is eminently 
necessary in exceptional cases in which not only 
these, but sodium and other soluble salts abound in 
the only obtainable supply. Ordinarily, however, 
it is too expensive, costing much more than a cent 
per gallon. Of the salts which render water hard, 
carbonate of lime is most common. This is kept dis- 
solved by the carbonic acid present, which is driven 
off by boiling. Boiling, therefore, causes most of 
the carbonate to become insoluble, and to separate 
as a precipitate. This portion thus removable is 
called ** temporary hardness. " The same result is 
obtainable by adding water-slaked quicklime in 
proper amount, estimated according to the degree of 
temporary hardness. If there be 20 parts of this 
to 100,000 parts of water (which is more than can 
well be tolerated for all domestic and industrial 
purposes), about 11.2 parts of fresh quicklime are re- 
(juired to counteract it. This, the so-called Clark 
process, can be accomplished rapidly by the aid of 



57 

several styles of more or less effective apparatus that 
agitate and aerify the treated water. Ordinary 
mechanical filters are also introduced into the proc- 
ess, or water may be allowed to settle in a tank for 
hours after the quicklime, wetted with four or five 
times its bulk of water and this still further diluted, is 
well diffused throughout the entire volume of the hard 
water. After a day or less of settling, the clear water 
may be drawn off, or the whole filtered. As above 
explained, the process reduces the percentage of bac- 
teria and organic matter also, as well as the greater 
part of the lime (and magnesia) carbonate. 

As a check upon the addition of an excess of lime, 
a solution of silver nitrite is recommended. If a little 
of this be added to a sample of the treated water and 
a brown precipitate results, it indicates that no more 
of the lime solution is needed. For each part of 
sulphate of lime and magnesia (''permanent hard- 
ness "), 2.86 parts of fresh sal -soda are required to 
counteract it. 

Springs and wells are fed from the vast quantity 
of water, usually very free from bacteria, that is 
present in permeable ground underlain by imper- 
meable strata. Artesian-well supplies have, in ad- 
dition, an impermeable stratum overlying the water- 
bearing strata. This supply is derived from rain 
and snow precipitated upon and sinking into the 
earth, perhaps a great distance away, and is termed 
ground-water. It flows along several feet a day, 
seeking its level toward streams and bodies of 
water. ^ 

When intercepted by rocky or other barriers which 

^ See The Am, Jour, of the Med. Sciences y December, 1890. 



58 

are impermeable, or nearly so, it tends to flow out 
as a spring, which, if favorably situated, is free from 
bacteria, as the water flows from the ground, be- 
cause of the slow flow through an immense mass of 
soil that furnishes the most perfect of filters. If 
into fine, porous soil, well situated, we drive a per- 
fectly clean tube, water similar to that of a pure 
spring is obtained. The supply remains free from 
bacteria so long as they are not introduced by con- 
tamination from above. Practically the same result, 
or nearly as good, can be obtained from a carefully 
dug well in a favorable spot, yet if a well be not 
properly curbed, covered, and cared for, dirt and 
bacteria are certain to enter it, and altogether too 
often prove the source of disease. Unclean buckets, 
failure to conduct off overflow (waste) water, filth 
carried in by rain, wind, general carelessness, and 
especially nearness of privies, barnyards, and other 
obvious sources of filth, account for the most fre- 
quent unsanitary impurity of well waters. Topo- 
graphic peculiarities, including flaws in the forma- 
tion into which the well is sunk, account for some 
contaminations. All germs of disease ought to and 
by due care can be prevented from entering. 

Wells may be regarded as large pipes surrounding 
an excavation in water-bearing ground and filled to 
a varying height with water from this cool source. 
If only unpiped holes in the ground, they are not 
hygienically complete. If, then, this pipe, covered 
on top, has an impervious wall tightly fitting in 
flawless soil and carried up at the top so that no 
surface washings can be introduced, unless filtered 
by at least eight or more feet of fine, close soil, 



59 

disease bacteria cannot enter. Then all water, how- 
ever it may vary chemically because of lime, iron, 
nitrates, chlorids, or what not else, enters only at 
the bottom through the perfect filter of the fine soil, 
and so is practically free from bacteria. The im- 
mense majority of observers agree upon this point, 
although occasionally some one comes forward with 
a statement to the contrary. An eminent biologist 
has recently reported finding bacteria in deep -driven 
well waters, but fails to explain that they, like those 
found by others under such circumstances, were due 
to contamination from above. Care should be taken 
during construction to see that the bottom of a dug 
well has no flaw of the rock or soil to allow bacteria 
to enter in that unusual way. The well being dug 
in a place free from danger of contamination of this 
sort, attention should be paid to the absolute pre- 
vention of direct pollution from the surface, which 
is the source of infection in almost all the cases in 
which well water causes typhoid and other water- 
borne disease. 

From the lowest part of the wall of the well -shaft, or 
at least from eight to ten feet below the surface of the 
ground, there should be a continuous, water-tight 
concreted casing, and this should rise from one to 
three feet above the ground. If concrete pipes, two 
or three feet in diameter, be employed for the pur- 
pose they must be well cemented at the joints and 
loam carefully rammed in around the pipe up to the 
level of the joint before setting and fixing the next 
length, all being supported so as not to settle. The 
top should be protected by a locked or otherwise 
fastened cover, asphalted, so as to keep out surface 



6o 

or other waste water, insects, and other things. 
Lead should not be employed for the purpose. The 
ground should be graded so as to slope away from 
the curb. Pails and buckets are not to be used. If 
a pump be placed over the shaft, provision should 
exist for conducting off waste water. It is much 
better that the pump be away from the well and the 
small (two-inch) pipe used for the pump-connection be 
carried through the casing at least several feet below 
the surface of the ground by a hole tightly cemented 
or asphalted about its periphery so that it allows no 
leakage into the well. The pipe may usually with 
impunity be of lead if the water be quite hard, but 
well-tinned brass or copper is better. Tin pipe, 
practically free from lead, could until recently be 
bought for a little more than 1 6 cents per pound, 
and ought to be used more frecjuently. The com- 
mon form of iron suction-pump is preferable to the 
endless chain and wooden pumps which are used over 
very many wells. 

As above indicated, wells which are properly con- 
structed and located many feet away from privies, barn- 
yards, wash-water, and other sources of filth that in no 
case drain toward the shaft, afford healthful water, pref- 
erable to that from rain, storage, or any ordinary surface 
supply. It is usually from surface wash entering, 
owing to the lack of an impermeable curb, casing, 
and cover, that germs of disease enter a well and 
cause typhoid fever and other water-borne diseases. 
In this sort of pollution is included filth brought 
upon the boots of users to the cover of the well or 
to the ground near it and washed in, as by waste- 
water entering through a defective cover, or carried 



6i 

in by rain or by leaching in over the edge with 
puddle- water or drainings. Water from shallow, 
unprotected wells, which are not grossly faulty in 
this regard, is likely to be quite clear and taste 
better than that from deep waters which contain less 
oxygen, but which are usually much freer from bac- 
teria. That deep protected wells receive much less 
of the only slightly changed organic and other 
chemic constituents of the soil which are carried 
down by heavy rainfall does not count for much 
hygienically, for it is particularly putrefactive and 
specific disease bacteria which cause shallow, unpro- 
tected wells to be dangerous. 

Palmer of the University of Illinois, in a recent 
chemical report, has added to our great volume of 
instructive statistics concerning Illinois and other 
States, which show the unwholesomeness of water from 
such mere holes in the ground. He emphasizes the 
familiar fact that deeper wells in the glacial drift of 
the prairie regions are much less productive of disease, 
although the water is often unpalatable, not only be- 
cause of lack of oxygen but also owing to presence 
of marsh-gas ( due to decomposition of vegetable 
matters) and perhaps of sulphuretted hydrogen. Iron 
often adds to the unpleasant taste and appearance of 
such water, so that the senses are likely to mislead 
people to prefer an inferior product. This explains 
the slowness with which the unintelligent learn that 
it is of paramount importance to guard wells against 
surface contamination. In prairie regions where the 
ground lies perfectly level, the house well ought to 
be a hundred feet away from any privy or other filth, 
and slops and kitchen waste should be carried far 



from the locality of the well. When the ground 
slopes, it must be borne in mind that water seeks a 
lower level, hence the well should be so situated that 
the natural flow of the ground-water current is from 
the well toward any privy or other source of con- 
tamination and not in the reverse direction. The 
ground must be studied to prevent the location of all 
three of these near together in one common small 
basin or sink formed by impermeable underground 
barriers. Mere vicinity of well and house is not 
necessarily dangerous if no privy or drain be near, 
and if no slops or other filth or surface drainage can 
leak into or upon the ground near by. The writer 
has made bacteriologic tests of carefully constructed 
wells adjoining, as also underneath houses with thor- 
oughly cemented cellars, and found the water ex- 
ceedingly pure and likely to remain so as long as no 
contamination entered; but it is invariably best to 
have such a source of drinking-water so situated that 
any one can always see that no carelessness ever in- 
troduces contamination. 

Local inspection and survey always furnish the 
most important means of determining the absence of 
contaminating influence. The shaft of the well is 
best examined if illuminated, and when emptied the 
lowest parts can be inspected and flaws carefully 
searched for. Chemic and especially bacteriologic 
tests if repeated several times under varying condi- 
tions are of considerable value in this respect. For 
rough practical purposes an inexpensive test, substi- 
tuting the lithium test, is afforded by throwing a 
suitable amount of saprol or some intense anilin red, 
dissolved in water, upon a neighboring filth pile or 



63 

in the kitchen slop -water suspected of polluting the 
well. Common salt can be made to serve the pur- 
pose. Saprol is a low-priced coal-tar disinfectant of 
a very marked odor and taste suggesting illuminating 
gas or naphtholin and, if it penetrates to the water 
of the well, betrays its presence in a sample taken. 
To the taste it is perceptible when only one part is 
present in two million parts of water. The odor is 
evident when twice as much is present in the 
sample. If none of this substance be detected in the 
well after liberal application of it to the suspected 
filth, it is presumably safe to use the water. If saprol 
enters the well, further and more precise examination 
should be undertaken to determine the defect. 
When the water of a well is exposed to pollution 
which cannot be prevented, and the trouble is not 
remediable by the introduction of an impervious 
curbing and the other safeguards above recommended, 
the safest way is to fill up the well and make a proper 
new one in a good location. Koch's recommenda- 
tion to convert a faulty well into a driven well by 
filling it up with sand and sinking a pipe in it is not 
practicable in most cases. At any rate, if this ex- 
pedient be tried, thorough disinfection and cleansing 
should precede its employment. 

In towns and wherever else it is available, steam 
driven in for two or three hours at a pressure of from 
two to four atmospheres is an efficient disinfectant. 
The leather valve of the pump is to be treated apart 
by soaking it for an hour in a i to looo solution of 
corrosive sublimate. After steaming has ceased, the 
water should remain in the closed well for nearly a 
day and then be removed together with the dead 



64 

frogs and other organic matter which, by subsequent 
decomposition, would become offensive, as would be 
the case after the use of lime or other chemicals. One 
portion of freshly slaked lime thrown in for every 
fifty parts of water proves an efficient disinfectant. 
It is cheaper than bromin or permanganate of potash 
and more satisfactory than the addition of one per 
cent, of sulphuric acid. Steam acts upon the entire 
well, while, when chemicals are added to the water, 
the walls have to be treated in addition. For artesian 
and other driven wells, Neisser recommends that 
steam be driven to the bottom of the tube by means 
of a small pipe sunk nearly to the bottom. Another 
pipe may also be introduced for the purpose of draw- 
ing out the water at the same time. 

Ground water derived from carefully located gang 
wells and used promptly without being stored is usu- 
ally very free from bacteria and other organic life. 
This is especially true when the territory which the 
driven pipes drain is reserved for a park and guarded 
from contamination. Such sources may in the end 
prove to yield less than the expected amount. If 
then, undue suction be employed to augment the 
flow of water, the results are unsatisfactory, and 
underground defects and partial stoppages of the 
pipes may be produced. When locating such a plant, 
bacteriologic tests are indispensable for preliminary 
determination of the quality of the water derived 
through sterile experimental tubes, samples being 
taken at different stages of the pumping. A chemic 
examination is also necessary for learning the amount 
of iron, hardness, chlorids, and organic matter. • 

Artesian wells usually yield water of great purity 



65 

as regards bacteria, yet tests of the water from these 
deep borings sometimes show numerous bacteria 
other than those introduced from above ground 
through carelessness. The character of those found 
aids in determining their source. There are cases 
in which various forms of water- life have been car- 
ried long distances through coarse strata, and the 
possibility of contamination from the remote outcrop 
of the water-bearing strata, must be borne in mind. 
In those regions in which these deep wells have most 
frequently been bored, the distant outcrop is usually so 
situated as to preclude the likelihood of infection en- 
tering the water there. Considerable salt occurs in 
artesian well water in some districts.^ With all their 
usual excellence as regards absence of bacteria, deep 
ground waters, even if the yield continues ample, 
proves at times too hard and contain too much iron 
to make them satisfactory, unless they undergo treat- 
ment. Moreover, the sieves introduced on the lower 
end of the driven tubes for the purpose of keeping 
out sand, tend to become incrusted and thus greatly 
retard the flow. To remedy this obstruction, hydro- 
chloric acid, diluted with nine times its volume of 
water, may be carried down to the bottom of the tube 
through a small pipe, and allowed to exercise its sol- 
vent action for nearly fifteen minutes, after which 
time it is pumped out. 

Iron, so' common in water from deep wells, may 
in some cases occur as phosphates of iron in part, 
but is usually in the form of free ferrocarbonate dis- 



» An excellent monograph on " The Artesian Wells of Dakota," 
by Gilbert, is contained in the report xviii, 2cl part, of the U.S,Geo* 
logical Survey. 



66 

solved in the carbonic acid of the water. If this be 
driven off in the presence of air, or, if the iron water 
merely comes briefly into contact with air, the iron 
carbonate is thereby changed into insoluble ferric 
compounds; consequently, according to the amount 
of iron present, and the length of time that the oxy- 
gen of the air has worked, the water becomes opal- 
escent, then gradually yellowish -brown, because of 
the innumerable minute particles precipitated, and 
which can be filtered out, leaving the. water clear. 
Earthly alkalies present favor the separation of the 
iron. Any contrivance, such as a spray or an aerified 
filter, which causes the water to be exposed to the 
action of air will remove nearly all of the iron from 
a water. 

Filtration, through from seven to ten feet of pieces 
of coke, gravel, or broken rock, one millimeter in size, 
with the utmost possible exposure to air, has been 
found in German experience to remove iron success- 
fully. The slimy precipitate collects and the upper 
layer of the filter has to be cleaned every fortnight 
or so, and the entire mass needs renovating, or re- 
newing every few months. Dunbar^ has indicated a 
suitable mechanic sand-filter for use on a small scale. 
The Krohnke process for removing iron calls for the 
addition of ten parts of ferrosulphate of iron to one 
million parts of the iron water. Then from fifty to 
a hundred parts of lime (with water) are stirred in 
and the whole allowed to stand for twenty minutes 
before filtering. Hardness, as well as organic con- 
stituents, is also greatly lessened by this process, which 
at the same time separates most of the bacteria. 

» Ztschr, fUr Hyg^^ xxii, 68, 



67 

Stcckel's process, used in connection with an ordi- 
nary well, consists in packing a layer of lime, or, 
preferably, tribasic phosphate of lime, on the bottom 
and sides of the well, and fixing it in place so that 
the entering ground water passes through the lime, 
which remains efficient for years without removal. 
It greatly increases the hardness of the water, but re- 
moves iron. 

For domestic use a bone-charcoal filter will attract 
and separate iron, lead, or other metallic impurities, 
present in small amount in water which flows slowly 
through it. Even the ordinary carbon cylinders sold 
for filtering have some value in this way, although it 
must be remembered that they have the defects of all 
ordinary filters. If such filters become clogged by 
the iron slime, they may be renovated by soaking 
them in dilute hydrochloric acid. If such a filter be 
used as a makeshift, the cylinder may be covered 
with coarse sand, and the water aerified by dropping 
upon it from a height. Whether any of these or a 
mechanical filter be used for deferrating water, the 
main thing is to see that abundant exposure to air 
takes place. 

Rain water, being precipitated moisture of the at- 
mosphere, is very pure as regards the presence of 
lime salts and other things taken up by surface and 
ground waters. As ordinarily obtained from a roof, 
through a gutter and pipe, the first portion that falls 
takes up not only the soot, dust, and grass which may 
be in the air, especially of cities, but the washings of 
the roof and gutters as well. It is best that the first 
portion which falls be allowed to flow off, carrying 
leaves, bird droppings, and other dirt, the following 



68 

portion then proving much purer. Automatic con- 
trivances exist to effect this separation of the unclean 
first flow, but are not likely to prove popular. They 
are apt to work badly when leaves or other coarse 
dirt are washed in. When rain water is of necessity 
depended upon for a supply, the water-conductor 
should be kept shifted so that the water running down 
after rain has begun to fall will be diverted until 
after the dirt has been washed off, then the following 
water may very easily be guided into the cistern. In 
many cases this water contains very many bacteria, 
which thrive in the darkness and abundant organic 
matter. Boiling renders it fit to drink. The cis- 
terns or tanks employed to hold water become foul if 
of wood, and should never have lead in any part, 
even for piping or cement. Bricks, well coated with 
hydraulic cement, are approved for cistern construc- 
tion. Reservoirs of any sort need cleaning from time 
to time. Chemicals, such as above indicated for 
porcelain filter-tubes, serve the purpose of disinfect- 
ing these also. Steam is excellent when it can be 
applied. 

SUMMARY. 

Water may be very clear and palatable, yet at the 
same time may chance to be impure. As precipitated 
by condensation of vapor it is pure. Precipitated 
from the clouds and sinking into the earth's surface 
and other natural reservoirs it remains wholesome so 
long as bacteria capable of inducing disease do not 
enter it. Bacteria present in almost all natural sur- 
face waters remote from habitation are presumably 
harmless. Harmful micro-organisms tend to perish 
or become innocuous in course of time after entering 



69 

ordinary water, yet complex conditions may make the 
operation of this principle uncertain. If harmful 
bacteria are known to be present more or less regu- 
larly in a given supply, or if the entrance of sewage 
or other obvious pollution near an intake make its 
presence probable, it is most necessary to employ 
adequate filtration or other means of purifying water 
taken from that place. Distilled water is whole- 
some if it contain no lead. The receptacles in which 
this and various originally pure bottled waters are 
stored and transported are often bacteriologically un- 
clean. Boiling, even for a moment, destroys the 
causative germs of typhoid fever, cholera, and other 
diseases known to be likely to occur in polluted wa- 
ters. Ordinary small filters are usually quite inade- 
quate to hold back disease bacteria or any others. 
All of them allow bacteria to multiply in the filtering 
substance and to pass through into the filtered water. 
Reported cases show that ordinary filters can harbor 
disease germs and cause disease. Small filters of 
very dense texture, such as the best porous porcelain 
and also the material used in the Berkefeld filter, can 
hold back all micro-organisms and thus yield abso- 
lutely sterile water for a limited number of days if in- 
telligently cared for. None of these are permanently 
* * germ-proof. ' * 

The only permanently germ -proof filter is the nat- 
ural one formed by immense quantities of fine porous 
earth, through which ''grouiid water '* slowly flows. 
Filtration by well -managed large gravity filter-beds 
of fine sand approximates this natural process in effi- 
ciency, being capable of rendering the water practi- 
cally pure. Mechanical filters can accomplish the same 



70 

result, but are less likely to be maintained constantly 
in a condition of the highest efficiency. 

Ground water, uncontaminated, is hygienically 
pure, under the conditions indicated above. In very 
many cases it is the best available source of supply. 
Free from bacteria as it issues from favorable ground, it 
frequently becomes contaminated because of careless- 
ness. An outlay of effort and of money, usually 
much less than would be required for securing a pure 
surface supply, will eliminate the defect, which is 
commonly a local one. Open, shallow wells, un- 
protected from direct pollution by surface washings 
from privies and other filth, are a notorious source 
of pollution by, and dissemination of, the germs of 
typhoid fever and other diarrheal diseases, including 
cholera whenever and wherever that occurs. If prop- 
erly located, and provided with a continuous water- 
tight casing and curb tightly covered, and if it has a 
good pump which does not necessitate an aperture 
above ground, even a shallow well can furnish whole- 
some water.