m m M.M..m

.t-it^,'

i' f

.■■'M''l

ft>:. S?!^.;

^" i-rtr#

't^'frf

%■ •«. #

a. '41;

®(|E ^. p. JIUI pfarmrg

J^orll) Olarolina ^t^ite College

QH51 _35

lJcI^C L^tiw

|lQiVlr'27

bMar35

24Wlay'51 X

140ct52j

BIOLOGICAL STUDIES

:a y£

BI()LOGIC7\L STUDIES

•UPILS OF

M SEDGWICK

H (

-^^^^^^^^^^'^'^•■'■■'^■^iliiililiiil'Mw^ri'r^hnrnr'r'*'*^^ iTrtrxn^nf^'j^j*'t^j'i^^^fn''^^'^*'^-^'^*--'

\M

^.^,^

BIOLOGICAL STUDIES

BY THE PUPILS OF

WILLIAM THOMPSON SEDGWICK

Published in Commemoration

OF THE Twenty-fifth Anniversary

OF HIS Doctorate

BOSTON

JUNE, 1906

PRINTED AT

THE UNIVERSITY OF CHICAGO PKESS

1906

THIS BOOK IS DEDICATED BY HIS PUPILS TO

WILLIAM THOMPSON SEDGWICK

TO EXPRESS THEIR REGARD AND ADMIRATION FOR HIM AS A FRIEND, TEACHER, INVESTIGATOR, AND PUBLIC-SPIRITED CITIZEN, AND ALSO TO AFFIRM THEIR LOYALTY TO THE IDEALS FOR WHICH HE HAS ALWAYS STOOD.

'T

A C-

TABLE OF CONTENTS.

PAGE

Calkins, Gary N. Paramecium aurelia and Paramecium caudatum - i-io

Dyar, Harrison G. The Life-History of a Cochlidian Moth —

Adoneta bicaudata Dyar 11-19

Fuller, George W. Experimental Methods as Applied to Water-

and Sewage-Works for Large Communities . - - - 20-35 ,

Leighton, Marshall O. The Futility of a Sanitary Water Analysis

as a Test of Potability 36-53

Whipple, George C. The Value of Pure Water - . - - 54-80

Mathews, A. P. A Contribution to the General Principles of the

Pharmacodynamics of Salts and Drugs 81-118

Stiles, Percy G., and Milliken, Carl S. An Instance of the

Apparent Antitoxic Action of Salts 119-123

Jordan, Edwin O. Experiments with Bacterial Enzymes - - - 124-145

.^WiNSLOW, C.-E. A., and Rogers, Anne F. A Statistical Study of

Generic Characters in the Coccacea; 146-207

^Prescott, Samuel C. The Occurrence of Organisms of Sanitary Sig- nificance on Grains -.- 208-222

^^Gage, Stephen DeM. A Study of the Numbers of Bacteria Develop- ing at Different Temperatures and of the Ratios between Such Numbers with Reference to Their Significance in the Interpreta- tion of Water Analysis ....---- 223-257

' WiNSLOW, C.-E. A., AND Lochridge, E. E. The Toxic Effect of Cer- tain Acids upon Typhoid and Colon Bacilli in Relation to the Degree of Their Dissociation -.--..- 258-282

Phelps, Earle B. The Inhibiting Effect of Certain Organic Sub- stances upon the Germicidal Action of Copper Sulphate - - 283-291

Jackson, Daniel D. A New Solution for the Presumptive Test for

Bacillus coli . . - -- 292-299

Ayers, S. Henry. B. coli in Market Oysters 300-303

Wadsworth, Augustus. Studies on Simple and Differential Methods

of Staining Encapsulated Pneumococci in Smear and Section - - 304-312

Kendall, Arthur I. An Apparatus for Testing the Value of Fumi- gating Agents 3i3~320

vii

viii Table of Contents

PAGE

Hough, Theodore, and Ham, Clara E. The Effect of Subcutaneous Injections of Water, Ringer's Fluid, and Ten Per Cent Solution of Ethyl Alcohol upon the Course of Fatigue in the Excised Mus- cles of the Frog - - . 321-326

RiCKARDS, Burt R. Notes on a Case of Apparent Pulmonary Tu- berculosis Associated with the Constant Presence of Diphtheria- Like Organisms in the Sputum ------- 327-329

PARAMECIUM AURELIA AND PARAMECIUM CAU- DA TUM*

Gary N. Calkins, Ph.D.

At the present time, when the subject of mutations and species is discussed on every hand, and when every eye is keenly on the alert for new evidence among animals and plants, a sudden trans- formation of one known species into another known species is of interest. Such an incident has recently come under my observa- tion; a Paramecium caudatum became P. aurelia, and remained so for about 45 generations, when it reverted to P. caudatum. Apart from the facts of the change, which in itself is of obvious importance from the standpoint of cellular biology, the essential question to consider is whether these two species are sufficiently well defined to justify their separation. If not, then the experiments and the changes indicated have less bearing on the general problem of mu- tation than upon the problems of cell physiology. If they are suffi- ciently distinct, then we have in this incident an interesting case of mu- tation. I personally believe that the slight differences that distinguish the two types of Paramecium are not of specific value, and hold^that P. caudatum should be regarded as a mere variant of P. aurelia. Paramecium aurelia was the name given by Muller, in his general work on the Protozoa in 1773, to the ciliated organism which had been known as the "slipper animalcule." Several different species of Paramecium were created by Ehrenberg in 1838, and described in his work on the Injusionsthierchen. Most of these have been sifted out into other genera, and only three have remained, P. bur- saria, P. aurelia, and P. caudatum. Paramecium caudatum and P. aurelia have been united into a single species by the majority of observers subsequent to Ehrenberg, on the ground that the differ- ences upon which Ehrenberg had based his species were inadequate. The number of species was thus reduced to two, and the names used were P. aurelia and P. bursaria, the former having been given originally by Muller. Maupas, however, in 1887-89, and R. Hert-

* Received for publicatioQ March 17, igo6.

jqtOrERIT IMAM

K. C State C»ll*f

2 Gary N. Calkins

wig in 1889, discovered a difference in the two forms which appeared to have specific value, and since then the two species in question, caudaiiim and aurelia, have been generally accepted as "good" species.

Paramecium aurelia, according to Maupas, differs from P. cau- datum in the following points: It is smaller (70 to 290 ft, as against 120 to 325 /i in P. caudatum); its posterior end is rounded, while P. caudatum has an attenuated end (hence caudatum); it has two small micronuclei (3 to 5 /a in diameter), while P. caudatum has but one (8 to 10 /tt); in conjugation its macronucleus becomes "rib- bon "-shaped at an earlier period than in P. caudatum; and after conjugation its cleavage nucleus gives rise to four corpuscles, whereas in P. caudatum there are eight. In deciding which of these forms to call caudatum and which aurelia, Maupas could not determine which type Miiller had seen, and went back therefore only to Ehrenberg, who' in naming P. caudatum had noted the attenuated posterior end. Hence it turns out that the more common form of Para- mecium has become widely known as P. caudatum, while the less common form bears the original name P. aurelia. If the two are only variants of the same species, it follow^s from the rules of zo- ological nomenclature that the common and well-known name Para- mecium caudatum must be given up and P. aurelia substituted. That this must be the case follows, as I believe, from the observa- tions here described. In the following description the names P. caudatum and P. aurelia will be used for those variants of the organ- isms which agree with Maupas' specific characteristics.

On March 11, 1905, four pairs of conjugating Paramecium caudatum were isolated from a culture that had been running for some weeks in the laboratory. Each pair was confined in a hollow ground slide in a medium of hay infusion made the previous day by boihng a small quantity of hay in tap water. The usual period of conjugation is from 18 to 24 hours, and by the following day all of the pairs had separated, and the different individuals were swimming about freely in the hay infusion as apparently normal ex-conjugants. Each individual was isolated and fed on hay infu- sion, and each became the progenitor of a more or less extended Hne of Paramecia, the method followed being the same as that described

o

t— «

H

<

< a. u

w

H <

K

H

O

OS Q

1/2

H

< o

S"

o u

I X

u o

CO

o

in

>

t— < Q

	00	■^	j^	Ov
	M	PD	'tf-	■w
r*	Tf	o	»>.
	'-'	rn	'^	■»r
o	-*	o.	-o
		f^	TO	■*
u^	't	00	VI
	•^	<ri	^	n
■^	WD	t^	(^
	M	(^	fO	•»
fO	M	iri	^
		fO	rO	•»•
«	O	^	0
	*^	(N	pn	■^
M	00	^	o.
A	M	N	r'l	fO
O	r>.	Ov	!>.
(H	N	«	m
g

■<	o-	vO	00
00	UD	r*.	in
	N	N	to
(>•	-t	o	-t
	W	«	"^
*o		^ N
»n	M	rO	„
	N	N	t«3
^	O	N	o
	w
fO	O	„	00
	M	c*	w
N	Ov	„	1^
		w	(S
	00	O	>o
		M	p)	w
	H	00	O	■"I-
	ro	"	■"	N
O	00	00	«
	rn	■^	^	fN
o	r^	t^	o
	M	M	**	w
00	o	o	o-
	c*	M	M	^
r^	"■*	u-i	o
	w	M	M	»H
so	f^	f*^	00
	r4		-s
		OJ
	ir:	N	MT3	t^
	N	"^		M
■^	^		-o
	(N	«3	M
	O
	(^	o-o	O rri	ir,
	(N
o
W	0> M	00 ^	•»
<	N
S

	M	00 w	MM r>. f^	fO
	W
0	r* M	MM \0 fO	<N
	w
o	ir, .-t	MM U^ N	o>

00	rj- M	MM -^t <N	t^

r^	« •-«	MM ff) N	xn
		-13
	<u
	NO	IN -H "O	"-• M P) p)	■O
	M
lO	MM M	MM P) M	•*
	■^	tn
	0
	Tj-	MM M ^^	MM MM
	•^
f^	MM MM	MM MM	„

		<ca OQ	w'ui OK	X
		^^— ' ^— ^—	' — .--^ • — ., — '

	^	O			^ 1
		o	O	o.
f^	00	CI
		NO	O^	o.
		s	o.
P<	:§	00	&
0	»n	s	00 00
o>	»^	ID	^
	"^		00	00
00	N	■<t	»D
	■"		00	00
t^	^	<*5	fo
	•^		00	00
<?	o.	M	^
	M	in	00	00
u^	r^	a	o
	•^		r^	00
•^	m	f^	00
	•^		f^	r^
•^	ID	•O	r*
<	•^	lO	f^	r*.
«	ro	^	r*
		»0	r^	r*
„	M	re	»D
	M	lO	r^	r*
o	„	^	"•r
		ID	<^	r*
o>	Oi	O	N
		•^	r*
00	"S.	00	O
		>o	r^
	^".	t^	00
		^	>o	o
^O	1^	ID	NO
		>o	NO
ID	vO	■V	t
	■*	vO	NO
^	VO	rn	■««■
	•>»■	VO	NO
PD	»n	«	«
		>o	-o
N	ID	^	M
	■<r	>o	NO
-	ID	vg	o NO

	0	ID	o	0
	rn		»o	NO
0^	ID	r*	o
	Pi	■«■	ID	ID
00	ID	r»	r*
	P)	•<«•	m	IT)
r-	fD	»D	ir,
	M	Tf	lO	»D
O	M	PO	^
iJ	W	■>t	»D	»D
ID	0	„	^
N	■<r	ID	tD
Pi 0»

<	t	oo	O^	■^
	M	rn	•t	»D
t^	o	00	■^
	PI	fD	^	ID
N	»D	NO	^
	PI	rD	t	>D
^	•n	ID	r^
	PI	fD	■(T	\n
o	•»	rO	«
	PI	rD	■"I-	wi
Q	■>»■	N	O
		<o	■<t	u	1
		"dm UQ	tiiui OE	X
		' — .— ' ' — „-^	» V — ' ■ V '

4 Gary N. Calkins

in a previous publication. As in previous work, the number of divisions was recorded each day for each of the ex-conjugants. In addition to the ex-conjugants, one individual from the original culture which had not conjugated, was isolated at the same time to serve as a sort of control. This line was designated X; the others, A-B, C-D, E-F, G-H, etc.

The early history of the division rate for all of these forms is given in the accompanying table.

The interest in the present paper centers in the first pair, A-B, for it was in this line of cultures that the aurelia form appeared. B died after a few days, but A recovered from the shock of reor- ganization and soon began to divide at a slow rate. As is the cus- tom in such work, one of the daughter-cells from an early division was killed and stained, to determine if conjugation had been nor- mally completed. The preparation shows that, instead of reor- ganizing with one micronucleus characteristic of P. caudatum, this individual had two, but otherwise appeared normal. From time to time after this individuals from this line of cultures were fixed and stained, and these preparations give a good history of the nuclear conditions at different periods. Some of these specimens are shown in photographs on Plate i. Fig. i represents an individ- ual in the third generation after conjugation, with four micronuclei (already divided for the ensuing generation) and the as yet incom- plete macronucleus. Fig. 2 represents an individual in the 24th generation; Fig. 3, one in the 43d; Fig. 4, one in the 46th; Fig. 6 shows two micronuclei in the end stage of division, the daughter- halves are connected by a delicate thread not seen in the photograph ; Fig. 7, finally, represents an individual in the 220th generation after complete recovery of the P. caudatum condition. The change from the one form to the other occurred during the first two weeks in May. Figs. 3, 4, and 5 represent three individuals from the same ancestor in the 43d generation. The first two individuals have each two micronuclei, the third has only one. All of these micronuclei show a marked increase in size over those shown in Fig. 2 for example, representing an individual in the 24th generation.

During the month of May and until three months after the cul- ture was started, individuals appeared here and there with but one

Paramecium Aurelia and Paramecium Caudatum

5

micronucleus, while the majority of them killed at this time appeared with one of the micronuclei larger than the other. By the end of June none of the P. aurelia forms were to be found, and this culture, like the other cultures started at the same time (G and X), contained forms with only one micronucleus; Paramecium aurelia had become P. caudatum again. This occurred between the 45th and the 70th generations, and the effect of this change upon the vigor of the race is evident from the remarkable rise which the accompanying curve representing the relative division rate takes at this period (see curve).

2<

tt

SI 21! 19 18 17 It

"••^•"

a

(

1!

If

x.— -

.— — ,

«

,1

1

1 !

- M.

BOM-)

r*rmL

1 ?

Ma-,-

i

TUNI -

i

r»iv^

>

k >

1

UCiliT

, i

c

1

Curves showing one nearly complete cycle, and the variations in -vitality as meas- ured by the division rate for ex-conjugants of the A series (solid line) and the G series (broken line). The ordinates represent the average number of divisions in lo-day periods. The rise in the A curve, as indicated in the first lo days in June, marks the recovery of the uninucleate condition characteristic of Paramecium caudatum.

Maupas held that the two species of Paramecium can be readily distinguished by the characteristics given above. If we examine these characters in the light of the experience with cultures, we find that they cannot hold good. For example, the relative size of the

6 Gary N. Calkins

two forms which Maupas, and with him more recently Simpson, held to be a distinctive feature, loses all value when considered critically, while size relations in general are absolutely untrust- worthy in settling questions of species. The variations which a species of Paramecium passes through under different conditions of vitality are so great, and last for such long periods, that no infer- ence can be drawn from cell dimensions at any given time. Size depends, apparently, upon two factors — the relative vitality, and the rate of cell division — and these two may probably be merged into one, which may be called the potential of vitality. Parame- cium caudatum under different conditions shows wide variations in size. When taken directly from the natural habitat, where food is not overabundant, they are large, measuring on the average 315/14. The same forms cultivated on hay infusion, with its rich food content, multiply rapidly, and do not grow individually to the same size as the "wild" form. These measure on the average (18 individuals killed at different periods of the cycle) only 206 /*, with variations from 180 to 224 /x when the division rate was rela- tively high — two divisions per day. When the potential of vitality is nearly exhausted and the division rate is low, a similar small size is noticed; but at such a time it is obviously due to a different cause, probably a loss of metabolic energy. The same differences are noticed in P. aurelia at different periods of vitality, and the impossibility of considering size relations as of specific value is clearly estabhshed. During the first 45 generations of P. aurelia the division rate averaged only eight-tenths of a division per day, which is lower than that for the G and X lines, which averaged one and one-tenth divisions. Eighteen individuals killed at differ- ent periods of the culture were measured during these 45 generations, and the average length was 224 /a, with variations from 168 to 256 /A. After the loss of one of the micronuclei the ♦> division rate increased to the remarkable rate of 2 . 2 per day on the average for a period of four months, when vitality waned. During this period of rapid multiplication the size averaged only 178 ft with variations from 148 to 212 /i. At this period the organ- isms in culture would have been identified by any microscopist as P. caudatum, although more than 40 /i, on the average, smaller

Paramecium Aurelia and Paramecium Caudatum 7

than the one which would be classed as P, aurelia, while the latter, in turn, measured 90 /* less, on the average, than wild P. caudatum. It is quite apparent, therefore, that size cannot be taken as a diag- nostic character in the present case; and this is, after all, only the application of a well-known principle in zoological taxonomy.

The pointed condition of the posterior end, also, in P. caudatum is likewise transitory, and may or may not be present in forms which agree in all other characters. Pointed Paramecium, if isolated, and the descendants watched for four or five generations, as I have done, will lose this characteristic and will become rounded and blunt at the posterior end (cf. Figs. 3, 4, and 7).

The sluggishness which Simpson advanced as a specific char- acter of P. aurelia is purely a physiological condition, depending upon the vitality at a given time, and is as much characteristic of P. caudatum as of P. aurelia.

The breaking-up of the macronucleus at an earlier period in conjugation, which Maupas considered a diagnostic feature of P. aurelia, may also be due to physiological conditions. I cannot write definitely on this point, as I have had no experience with con- jugating forms of P. aurelia. By itself it would not constitute a diagnostic characteristic of sufficient value to determine a species. So, too, the other characteristic feature of conjugating forms, named by Maupas, the number of corpuscles into which the fertilization micronucleus divides, would be dependent upon the number of micronuclei present, and would amount to the same thing in either case, if each of the two micronuclei of P. aurelia forms four micro- nuclei, as Maupas describes. The experience which I have de- scribed above of the presence of two micronuclei after conjugation in forms which had only one before, indicates that eight corpuscles characteristic of P. caudatum were also formed here, but were resolved into a binucleate instead of a uninucleate condition. This characteristic, therefore, cannot be termed specific.

Apart from the purely physiological characteristics which have little or no value in classification, there remains only the one specific feature to justify the attempt to separate P. aurelia and P. caudatum, viz., the presence of two and one micronuclei respectively. My experiments show that this, too, is inadequate, for P. caudatum

8 Gary N. Galkins

may become P. aurelia, and P. aurelia may in time lapse again into P. caudatiim. It is to be inferred from this that the various forms which have been described as P. aurelia are in reahty only "sports" of P. caudaium, and if such "sports" are unable to keep up by inheritance the characteristic structural differences which distinguish them from the ancestral form, although this is main- tained for the long period of 45 generations, they cannot be considered a "good species." On the other hand, R. Hertwig gives some evidence, not conclusive, however, to show that P. aure- lia, after conjugation, reorganizes with two micronuclei.

It must be admitted that one experience of this kind may be insufficient to throw out a species that has appeared to be so well established. It may be that my observation was made on a chance abnormality which paralleled P. aurelia, and that the real P. aurelia retains its integrity as a species. Personally, however, I do not believe it, and am reasonably confident that such abnormalities may be of frequent enough occurrence in nature to account for the numerous descriptions of P. aurelia that have been given. Forty generations is a long series for an abnormality to be transmitted, and the number of individuals represented by 2 to the 40th power (as my culture represents), allowing for natural loss through enemies, etc., would provide enough specimens with this abnor- mality to justify the belief that it is normal. On the other hand, it cannot be stated that P. aurelia is a well-established species. It is relatively rare in nature; its specific character has been con- tested by such eminent authorities as BiitschH, Engelmann, Bal- biani. Stein, Koelhker, and Gruber; while Maupas and Hertwig succeeded in establishing it as a species only on the slender basis given above. My one experience with this culture is strong enough, as I believe, to reanimate the old skepticism, and to justify us in abandoning either P. aurelia or P. caudaium. The latter is the more recent name given by Ehrenberg, and according to the rules of priority must be replaced by Paramecium aurelia, the name applied by O. F. Mijller to the "slipper animalcule."

The physiological features presented by this experiment give some interesting data upon the vitality and nuclear relations. The curve shows that a sudden rise in vigor accompanied the return

Paramecium Aurelia and Paramecium Caudatum 9

to the normal through the loss of one of the two nuclei. Although it is difficult to estimate accurately the relative volumes of nucleus and cell body, some idea of the relationship can be given by meas- uring the two dimensions of the surfaces exposed. The surface of the macro- or micronucleus obtained by multiplying the two dimensions exposed, have a measurable relation to the total surface of the organism, and in a rough way this relationship represents the relative volumes. Measured in this way, it was found that in resting, vegetative cells of P. aurelia the relation on the average of both micronuclei together to the whole organism is as 1:717 (or of the single micronucleus to total body as 1:1,434). In dividing cells this proportion rises to the ratio of i : 608 (or for the single micronucleus as 1:1,216). In the P. caudatum stage, after the loss of one micronucleus, the ratio of the micronucleus to the entire cell is expressed by the ratio of 1:648 for resting cells and 1:440 for dividing cells.

The macronucleus is generally considered to be the primary agent in constructive processes of the cell, and therefore of the first importance in considering the growth and division energy. In the P. aurelia phase, when the division rate was low, the average relation of the macronucleus to the whole body was as i : 43, with wide variations from 1:16 to 1:68. In the P. caudatum phase, with one micronucleus, the proportion of the macronucleus to the entire cell fell to 1:50 on the average, with variations from i : 39 to 1:82. The difference is not great, and may well fall within the limits of experimental error, and it is reasonable to infer that the physiological differences which are evident between the aurelia and caudatum forms do not owe their origin to the difference in the mass of macronuclear material. On a priori grounds it is reasonable to expect the more rapid multiplication in forms with the greater proportion of micronuclear material, and this is borne out in the experiments for it appears from the curve that during the period of abnormal micronuclei, when the smaller amount of micronuclear material was distributed in two nuclei, the organism was laboring under abnormal physiological conditions and had a lower division rate than when the normal uninucleate conditions were restored.

lo Gary N. Calkins

EXPLANATION OF PLATE i.

M = macronu cleus . m=micronucleus.

Fig. I. — Ex-conjugant of the A series in the third generation after union. Four micronuclei can be seen to the right of the macronucleus, and, in addition, a rela- tively large unabsorbed fragment of the old macronucleus. The micronuclei are precociously divided for the fourth generation.

Fig. 2. — Paramecium aurelia in the 24th generation. The two micronuclei may be seen at the lower end of the macronucleus, one on each side.

Fig. 3. — Paramecium aurelia in the 43d generation. The two micronuclei are of relatively large size, and are dissociated from the macronucleus at the upper end of the animal (the two objects at the right of the macronucleus are foreign par- ticles deposited on the organism).

Fig. 4. — Paramecium aurelia in the 46th generation. The two micronuclei are at the upper end of the macronucleus, and cannot be seen distinctly.

Fig. 5. — Paramecium aurelia in the caudatum phase at the 43d generation after union. The single micronucleus may be seen at the extreme right end of the. nuclear material. It is noticeably larger than the micronuclei of the aurelia phase (compare Fig. 2). This specimen is one of the same lot as that represented by Fig. 3.

Fig. 6. — Paramecium aurelia in the 32d generation vdth macronucleus and micronuclei in division. Three of the micronuclei are shown with pointed ends; the fourth is out of focus and does not show.

Fig. 7. — Paramecium caudatum phase of P. aurelia, in the 220th generation. The relatively large micronucleus is dissociated from the macronucleus. This soeci- men was killed when the division rate was high (end of July; cf. curve).

1 l^i K A. M J i.

:M.

. . . M

w

l-ic. 1.

\^^^S^^^^^

M

.»

Fig 2.

Tk;. 3.

IM... .,.

Fh-.. 5-

% %

Vu:. 6.

I'lG- y.

THE LIFE-HISTORY OF A COCHLIDIAN MOTH— .IDO- NETA BICAUDATA DYAR*

Harrison G. Dyar,

Custodian of Lepidoptera, U. S. National Museum, Washington, D. C.

The larvae of the Cochlididas, or slug caterpillars, are especi- ally interesting to the entomologist on account of their peculiar forms and considerable diversity of structure. The author has been interested in this family for twenty years, and has been so fortunate as to work out the life-histories of many of the species inhabiting the Atlantic coast region of North America, as well as a few foreign ones. The results have been pubHshed in a number of articles ;'"^^ so it is not necessary to repeat here the general struc- tural characters of these larvae. They are highly specialized forms of Tineoidea or Microlepidoptera; that is, the larvae are highly specialized. The moths have not shared in the complication of structure exhibited in their early stages. The family is spread throughout the world, and is doubtless of early origin in geologic time.

North of Florida and Texas, where peculiar forms occur, 26 species are at present known. Of these I have found the larvae of 20, and have been especially interested to secure the remain- ing six, if possible. It was the search for one of these, Monoleuca semijascia Walker, which resulted in the knowledge of the life- history of Adoneta hicaudata, though not successful in its immediate object. In 1898 and 1899 I made collections of larvae at Morris Plains, N. J., where it was said that Monoleuca had been taken many years ago, hoping that it might still be there. The larv£e were very small; but as nothing that looked to be the one sought appeared, they were turned over to Mr. L. H. Joutel, of New York. Later in the season Mr. Joutel called my attention to two larvae which were new to him that he had noticed among his stock, and which he thought were from among those I had given him, or else came from Long Island. The new larva was allied to Adoneta spinuloides Herrich-Schaeffer, and we thought it might be the Mono-

♦Received for publication October 5, 1905.

II

12 Harrison G. Dyar

leuca, since that genus is allied to Adoneta as adult moths. We were not then successful in rearing the new larva. It is referred to in print^'* as Monoleuca ( .?), as I then thought it might be. Sub- sequently, in arranging the collections at the National Museum at Washington, I found an inflated larva of this new form, together with a bred female adult, prepared by Mr. A. Koebele, formerly of the United States Department of Agriculture. The adult was not Monoleuca, but a light-colored form of Adoneta, which I supposed must be Packard's leucosigma, though I could not be sure that the new larva properly belonged to it, not finding Mr. Koebele 's notes. On the strength of this specimen, I listed leucosigma as a variety of spinuloides.^^ In 1903 and 1904 Mr. W. F. Fiske, of the Forestry Division of the Bureau of Entomology, was collecting at Tryon, N. C, and he turned up Monoleuca semijascia in numbers. At my request, Mr. Fiske secured eggs, but they all proved infertile. In the fall of 1904, therefore, I went to Tryon for the purpose of finding the Monoleuca larva. I was unsuccessful; but in the search obtained a number of larvee of the new form of Adoneta. In the same season a number of adults were collected at Plummer's Island in the Potomac River, near Washington, by Messrs. August Busck, H. S. Barber, and E. A. Schwarz, and it became evident that I had before me an undescribed species of Adoneta, which was not leucosigma Packard. I therefore published a description of it, 3° associating with it the new larva.

The larvae from Tryon resulted in several male and one female moths in July, 1905. The males emerged before the female, and, in flying in the cage, lost their legs, so that they quickly destroyed themselves by buzzing on the ground. When the female emerged, I had no living male left. I placed her in the cage at Rock Creek, near Massachusetts Avenue, in the District, but no male was at- tracted. The next night Mr. H. S. Barber kindly took the cage to Plummer's Island and remained there all night, but again with- out success. However, Mr. Barber took at light two males of this species, which he kept ahve in glass tubes. One had lost its legs and was useless, but the other was in good condition and was placed in the cage. Mating occurred on the third night after emer- gence of the female, and good eggs were secured.

Life-History of a Cochlidian Moth 13

Adoneta hicaiidata is remarkable for the lateness of emergence of the adults, late July and early August being the time of flight. One other species, Isochaeles beutenmuelleri Hy. Edwards, has similar habits and a similar distribution. Northern New Jersey and Staten Island N. Y.,^^ seem to be the limits of their range northward, and it is doubtful if they occur so far north in every season. Their lives as larvae are as long as those of any other species, namely about eight weeks, so that they are liable to be destroyed by early frosts before maturing. After spinning their hard cocoons on the ground or among dead leaves, they are immune to cold, though pupation does not occur till the following spring. The species is, of course, strictly single-brooded.

The eggs are laid in groups of two to ten, overlapping, and placed on the backs of the leaves. They are usually deposited in low places, young red oak bushes in not too dense woods being the best situation. None were found on the higher branches. At Tryon, in 1904, all the larvae taken were on small red oak trees. In 1905 I took a larva near Washington (Rosslyn, Va,), also on red oak, and a number at Tryon at the end of September on this tree, and also on other plants, white oak and Ceanothus. The little larvae, on hatching, separate, having no tendency to be gregarious. Their first business is to molt, which they do within two days of hatching, without having fed during the first instar. In the second stage feeding begins. The young larvae, up to the fifth or sixth stage, eat the lower epidermis and parenchyma of the leaf, cutti ig short, scattered, irregular channels of about the width of their bodies. They frequently change their feeding-place and pass readily from one leaf to another, but of course cannot leave the tree on which the eggs were deposited. When large enough, they begin to cat the whole leaf from the end or side inward, as the other species of Cochlidians do. The larvae observed by me passed nine stages, though they can, no doubt, mature in eight under more favorable conditions. I was obliged to carry my larvae to northern New York, where they were fed on yellow birch, oak not being available, and were subjected to the rigors of an Adirondack climate. When matured the larvae change color slightly and leave the tree to spin their cocoons among dead leaves on the ground, wherein they pass the winter in the prepupal stage.

14 Harrison G. Dyar

As compared with its nearest ally, Adoneta spinuloides H. S., the larva of A. hicaudata is narrower, less elevated at joint 5; the horns of joint 13 are long, predominant over those of joint 12, which exceed those of joint 10; the widenings of the dorsal purple band are all subequal, five widenings, each practically alike, no pushing out of the subdorsal horns of joints 6 and 7; caltropes reduced on the anterior side horns, joint 6, et seq., only well developed on joints 12 and 13; the color of the dorsal band is bluer purple than spinuloides, less reddish or maroon.

DESCRIPTION OF THE SEVERAL STAGES.

Egg. — Elliptical, colorless, shining, fiat, the surface faintly retic- ular. The yolk forms a slightly opaque central mass, leaving a transparent rim. On the leaf the eggs are shining and transparent, resembling drops of moisture. Their extreme flatness enables them to be laid overlapping hke shingles. Size, i. 4X0. 8X0. i mm. The development of the embryo can be easily observed in eggs laid on glass. The shell is thin and transparent, merely a delicate skin. It is impossible to detach the eggs without destroying them. They hatch in 10 days.

Stage I. — Elliptical, the dorsum flattened, the sides oblique, the venter fiat. Head small and weak, without hard chitinous covering; mandibles and palpi present, but reduced and weak; the mandibles with four equal blunt teeth incapable of feeding, colorless. A subdorsal and a lateral row of processes or "horns," one on a segment, the subdorsals on joints 3-13, the laterals on joints 3, 4, and 6-12. The subdorsals of joints 3, 4, 5, 8, 11, 12, and 13 are large, the others small; the laterals of joints 3 and 4 are large, the others smaller, but not as small as the small subdorsals. Each horn bears three setas, occasionally but two, slender, slightly clavate-tipped, smooth. The setae of joint 2 are not elevated on horns. Skin smooth. Colorless, whitish, without markings. On hatching, the larvae gathered in groups as they had been laid, and molted in two days. Length, 0.9 mm.

Stage II. — Head rounded, white, clypeus highly triangular, reaching about half-way to the vertex; eyes black; mandibles stout, brown-tipped. Elliptical, narrowed behind, dorsum nearly

Life-History of a Cochlidian Moth 15

level. Subdorsal horns of joints 3-5, 8, 11, 12 spherical, large, densely spined, those of joints 6, 7, 9, and 10 minute, with one tubercle and one spine only, and each approximated to the adjoining large horn; subdorsal horn of joint 13 small, with two or three spines. Lateral horns alike with three or four spines, except those of joints 3 and 4, which are larger and more rounded. Skin rather densely granular, smooth, without markings. The spines are large, with large tubercular bases and black tips. The detachable tip is nearly as long as the shaft of the spine and is not enlarged near the junction. These spines are presumably of the urticating type, as they have that structure, though they are probably too small to pierce the human skin, at least in this early stage. There is a row of fine hairs on the anterior edge of the hood (joint 2). Length, 0.9- 1 .6 mm.

Slage III. — Head elongate, higher than wide, whitish, the eyes in a round black spot, the mandible brown-tipped. Elongate ellip- tical, flattened, rounded squarish, narrowed behind more than in front, the dorsum flat. Subdorsal horns large, rounded, well spined, those of joints 6, 7, 9, and 10 represented by four or five spines on the skin, approximated to the neighboring horns. Lateral horns moderate, rounded, all elevated. Skin granular shagreened, the granules sharply conical over the dorsal region and separated by their own diameters or more, on the sides more flattened and irreg- ular till along the subventral edge they form a pattern resembling alligator skin. All whitish green, no marks. The depressed spaces (i) are faintly indicated, paired. The cores of the large subdorsal horns are slightly whitish. The spines on the horns have the detach- able tips relatively shorter than before, being not over one-third of the whole spine for the subdorsal horns. The lateral ones are less advanced, and have the structure of the subdorsals of the pre- vious stage. Length, i . 6-1 . 8 mm. This is apparently the inter- polated stage.

Stage IV. — Elongate, flattened, tapered behind, the ends trun- cate; the larva generally sits a little curved. Subdorsal horns rounded, large, subequal on joints 3, 4, 5, 8, 11, and 12, the pair on joint 13 small, not as yet produced. The short horns of joints 6, 7, 9, and 10 rounded, well spined, not so strongly appro .ximated

1 6 Harrison G. Dyar

to the neighboring horns as before. Lateral horns small, sub- equal. Skin densely granular shagreened. Depressed spaces (i) distinct, paired, a little whitish. Translucent greenish, a narrow white hne along the subdorsal ridge the whole length. Horns slightly yellowish-tinted, with a trace of vinous shading at the ends of the body. Later there appear more decided traces of color. A yellowish-white band in the subdorsal ridge is broken into six slightly obhque broad bars under the long horns, elsewhere faint; there is a fine hnear straight white dorsal line, broken, most distinct centrally, dividing the small white dots of depressed spaces (i). A slight purplish infiltration in the dorsal space between these marks. Length, 1.8-2.8 mm.

Stage V. — Elongate, flattened, tapered behind, the subdorsal horn of joint 13 about two-thirds as long as that of joint 12, the rest subequal, those of joints 6, 7, 9, and 10 very small. Lateral horns of joints 3 and 4 moderate, the rest small. Green, whitish along the sides; subdorsal horns yellow with a yellowish- white subdorsal band below them, narrowed in the spaces between to border the dorsal space, which is especially elliptically widened at the short horns of joints 6-7 and 9-10. A broken white dorsal line; depressed dots (i) greenish, impressed. A slight purple infiltration all along the dorsal space between the markings. Length, 2 . 8-4 . 1 mm.

Stage VI. — Horns of joint 13 distinctly elongated into tails> longer than the subdorsals of joint 12. EUiptical, the sides wide centrally and flattened, the dorsum long, straight, narrow. Horns moderate, the short subdorsals distinct, separate, about as large as the lateral horns, of which those of joints 3 and 4 are stouter, but hardly longer than the rest. Green, the subdorsal ridge broadly yellow, narrowed at the short horns, causing the dorsal space to expand there, apparently. A broken white dorsal line touching the white dots of depressed spaces (i). Dorsal space partly purple- filled. A broken yellow hne beneath the lateral horns. Subdorsal horn of joint 3 slightly red-tipped. Spines black-tipped; skin densely frosty granular; the lateral depressed spaces show faintly as large, concolorous, kidney-shaped hollows. Later the markings become more pronounced. Green, the subdorsal horns of joint 3

Life-History of a Cochlidian Moth 17

red, those of joints 4 and 5 red-tipped, the rest green. Dorsal marking expanded in the intersegments 3-4, 4-5, 5-8, 8-10, and very shghtly lo-ii and 11-12, the first and last green-filled, the rest dark purple, cut by the yellow dorsal line and the pale depressed spaces (i), broken at joint 8, broadly yellow-edged to the sub- dorsal horns. Horns with stinging spines, no caltropes. Skin sparsely clear granular. Length, 4.1-5.2 mm.

Stage VII. — As before. The subdorsal horns of joint 13 are long and tapered. The subdorsals of joints 3-5 are red, those of 12 and 13 sHghtly red-tipped; a greener fine within subdorsal ridge. Length, 5.2-6.9 mm.

Stage VIII. — Subdorsal horn of joint 3 a little longer than those of joints 4 and 5, but all now very short, those of joints 6-11 still shorter, of 12 a little longer, of 13 long and tail-Hke. Side horns all very short. The level narrow dorsum is yellow and opaque, the dorsal band of purple widens into five subequal patches, nearly divided at joint 8, where the purple is replaced by red and is Unear only. The last patch, on joints 11 and 12, is a Uttle smaller; joint 13 dorsally green. Horns bright red. Sides all clear green; a broken white line along the lateral ridge and irregular white marks in the lateral depressed spaces. Skin granular; spines pale; no caltropes, except perhaps two or three on the basal anterior green spot of the horn of joint 13. Dorsal impressed dots (i) greenish with white rings. A dorsal white line narrowly cuts the purple. Length, 6 . 9-9 . 3 mm.

Stage IX. — Long, rather narrow, quadrate, a little tapering behind. Dorsum broad, flat, not arched and scarcely higher at joint 5, yet a little so. Subdorsal ridge indicated by change in direction. Sides perpendicular or nearly so, the lateral space broad, continuous with the subventral space which is infolded in the middle. Subdorsal horns distinct, short, those of joints 3, 4, 5, and 12 mod- erate, those of joint 13 long, nearly three times as long as the ones on joint 12, the rest short, those of joints 8 and 11 a Httle larger than the others. Side horns short, sessile, wider than long, those of joints 3 and 4 a little longer than those of 6-12. Caltrope patches on the horns of joints 6-12 and on the base of the subdorsal horn of joint 13, large on joints 12 and 13, then progressively smaller

1 8 Harrison G. Dyar

till the horns of joints 6 and 7 have only a few or no caltropes. Skin finely clear granular except on the horns. No end spines. Dorsum yellow- or red-shaded, a purple band with white glandular dots and central dorsal hne much as in spinuloides, but of different shape. It widens between joints 3 and 4, 4 and 5, then moderately widens on joints 6 and 7, narrows to a sMght bordering of the white dorsal line over joint 8, widens behind the horns on 9 and 10, moderately widens between joints 11 and 12 and ends, joint 13 being green above. A bright red, diffuse, subdorsal band; all the subdorsal horns red. Below, a yellow stripe, narrowly red-edged, waved. Sides green, a row of yellow dashes along the lateral horns, green- edged above; yellow rings on spaces (4). A white Hne along the subventral edge. Stinging spines short, not numerous. Depressed spaces (i) and (2) represented by white dots, (i) paired and on joints 3-4 and 4-5 also double; depressed space (4) reniform, dis- tinct; sHght hollows subventrally ; spiracle of joint 5 moved up out of Hne. Length, 9.3-13 mm.

Cocoon. — Nearly spherical, hard, brown, one end opening Hke a circular lid on emergence, though there is no sign externally of the crevice of the lid. Spun with a slight veil against a leaf. Diam- eter, 5 mm; length, 7 mm.

Pupa. — With the characters of the family. DeHcate, thin- skinned with the members free. It emerges nearly out of the cocoon when the adult issues.

Food plants. — The red oak is the preferred plant, though the larvae will feed on almost any smooth-leaved tree. I found one on maple at Tryon, the others mostly on oak. In confinement my larvae fed on yellow birch, which they seemed to prefer to soft maple.

REFERENCES.

Dyar AND Morton. General account, Jour. N. Y. Ent. Soc, 1895, 3, pp. 145-51.

" " " Apoda y-inversa Pack., ibid., 1895, 3, pp. 151-57.

" " " Sibine stimulea Clem., ibid., 1896, 4, pp. 1-9.

Dyar. Euclea delphinii Boisd., Florida Form, ibid., 1896, 4, pp. 125-29.

" Tortricidia pallida H.-S., ibid., 1896, 4, pp. 167-72.

" Eulimacodes scapha Harr., ibid., 1896, 4, pp. 172-78.

" Phobetron pithecium Sm. & Abb., ibid., 1896, 4, pp. 178-84.

" Sisyrosea textula H.-S., ibid., 1896, 4, pp. 185-90.

" Tortricidia fasciola H.-S., ibid., 1897, 5, pp. 1-5.

Fig. I.

""'"■''■ ■'■'^

Fig.

_T-rf^^-^^'--''>'- "••-•^

■.■ ■„.,■. 11 1- '. ■i.'i-..,i^-nT^

Fig. 6.

Fig. 2.

Fig. 7.

Fig. 4.

-snr-"v-v :'.'.";' ;:,, ';.'.; ju. M^S3

Fig. 5.

Life-History of a Cochlidian Moth 19

10. Dyar. Adoneta spinidoides H.-S., ibid., 1897, 5, pp. 5-10.

11. " Enclea indetermina Boisd., ibid., 1897, 5, pp. 10-14.

12. " Euclea delphinii Boisd., ibid., 1897, 5, pp. 57-61.

13. " Parasa chloris H.-S., ibid., 1897, 5, pp. 61-66.

14. " Calybia slossoniae Pack., ibid., 1897, 5, pp. 121-26; 1898, 6, pp. 158-60.

15. " Apoda biguttata Pack., ibid., 1897, 5, pp. 167-70.

16. " Packardia geminata Pack., ibid., 1898, 6, pp. 1-5.

17. " Packardia elegans Pack., ibid., 1898, 6, pp. 5-9.

18. " Heterogenea fle.xuosa Grote, ibid., 1898, 6, pp. 94-98.

19. " Tortricidia testacea Pack., ibid., 1898, 6, pp. 151-55.

20. " Heterogenea shurtleffii Pack., ibid., 1898, 6, pp. 241-46,

21. " Natada nasoni Grote., ibid., 1899, 7, pp. 61-67.

22. " Cochlidion avellana Linn., ibid., 1899, 7, pp. 202-08.

23. " Summary and Conclusion, ibid., 1899, 7, pp. 234-53.

24. " Note on Doubtful larva, ibid., 1899, 7, 236 note.

25. " Sibine jusca Stoll, Ent. News, 1900, 11, pp. 517-26.

26. " Isochaetes heutenmuelleri Hy. Edw., Proc. Ent. Soc. Wash., 1901, 4, p. 300.

27. JouTEL. Note on Occurrence, Jour. N. Y. Ent. Soc, 1902, 10, p. 248.

28. Dyar. Catalogue of species. Bull. 52, U. S. N. M., 1903, pp. 354-58.

29. " Synonymy, Bull. 52, U. S. N. M., 1903, p. 355, No. 4085a.

30. " Description of new species, Jour. N. Y. Ent. Soc, 1904, 12, p. 43.

31. " Catalogue of species, Proc. U. S. N. M., 1905, 29, pp. 359-96.

EXPLANATION OF PLATE 2.

Fig. I. — The larva in Stage I, dorsal view.

Fig. 2. — One of the subdorsal horns of Stage I, the spines of the next stage appear- ing by transparency. The larva was about to molt. Fig. 3. — A spine of Stage II. Fig. 4. — The mature larva in position of feeding. Fig. 5. — A seta of the last stage. Fig. 6. — An urticating spine of the last stage. Fig. 7. — A caltrope spine of the last stage.

EXPERIMENTAL METHODS AS APPLIED TO WATER- AND SEWAGE-WORKS FOR LARGE COM- MUNITIES*

George W. Fuller.

That progress means advance in knowledge and the gradual transition of information from the unknown to the known is of course a truism. The manner by which progress is made in different lines of work varies widely. The experimental method as appHed to the teaching of science in educational institutions has been of great benefit, and the application in a more systematic manner than formerly of the so-called "cut and try" method of the early inventors has resulted in more substantial progress in many mechan- ical and industrial Hues. By the man of affairs this advance in knowledge is referred to as additional experience. It is not the purpose of this paper to attempt an analysis of the manner by which knowledge in general is advanced, but to refer somewhat briefly as a matter of record to the way in which there have been solved various sanitary problems which a quarter of a century ago, for financial or other reasons, seemed out of question.

It is fitting on this occasion that this topic should be touched upon, owing to the influence exerted upon this hne of work by the two institutions with which Professor Sedgwick has been most actively connected in recent years, viz., the Massachusetts State Board of Health and the Massachusetts Institute of Technology. The former has for years been the foremost among the state boards of health of America in leading local authorities to improve their water- and sewage- works in an efficient and economical manner. Its influence has extended, not only to other states, but also to numerous places throughout the civilized world. The Massachusetts Institute of Technology has exerted a less direct influence; but, as a training school for many who have taken a prominent part in improv- ing water- and sewage-works, this institution and a number of its

♦Received for publication March 24, 1906.

20

Experimental Methods in Water- and Sew age- Works 21

teaching stafif have achieved resuhs which will be more fully appre- ciated as the years go by.

While the experimental method so successfully applied in the student laboratory may in its way call for just as conscientious and dihgent effort as when applied to projects involving immense sums of money, yet the responsibility associated with the latter is far greater. It is necessary, in carrying out successfully these large practical problems, not only to draw correct conclusions from full representative premises of a complicated nature, but also to adjust the project to a reasonable business basis, to make it fairly well understood by non-technical officials and citizens, and to defend it from the obstructionists who, for political or selfish reasons, cross the pathway of nearly every large public enterprise. In meeting these requirements there has been called forth a series of efforts which are of great significance to the public from the sanitary and financial standpoints, and which form notable achievements in the field of apphed science.

benefits of improved sanitary works.

Improved water- and sewage-works of course do not explain by any means the entire improvement which for the past quarter of a century has been so characteristic of the sanitary conditions of a majority of the civilized communities of the world. But, illus- trative of the scope of such improvements under the guidance of wise sanitary authorities, it is of interest to point out the markedly decreased death-rates in Massachusetts from water-borne diseases, of which typhoid fever is the typical, but not the only one.

Death Rates per 100,000 Population from Typhoid Fever in the State of Massachusetts

BY Five-Year Periods from 1881 to Date.

Period

Rate

Period

Rate

iSSi-iSSs

41 46

34

1896-1900

igoi-1904

26

1886-1890

1891-1895

19

What is true of Massachusetts is in a general way true of many sections of both America and Europe where the population has become quite dense, and demands for water- and sewage-works of satisfactory character have pressed forward for attention in recent

22 George W. Fuller

years. No attempt will be made here to show the great sanitary benefit derived from other factors than improved water supply and sewerage, as this matter is clearly set forth in standard works upon sanitation and official reports from various quarters of health author- ities who deal with the accomphshments of the modern health officers.

It is sufficient here to present clearly to the reader the thought that modern sanitary science has greatly increased the comfort and safety of living. In various cities the reductions in death-rates have been far greater than as given above for the state of Massachusetts. This is especially true in cities where badly polluted water supplies have been replaced by improved supplies. Numerous instances of this sort are on record where annual typhoid fever deaths of 50 to 100 per 100,000 have been reduced to about 20. These cities include not only those now receiving upland waters from unpolluted sources, and ground water, but also those having filtered water from earlier but polluted sources; for example, Lawrence, Mass., Albany, N. Y,, York, Pa., and Lorain, Ohio. Intestinal diseases other than typhoid fever have been so reduced as to lower the general death-rate materially. For low-lying communities like New Orleans modern drainage has lessened notably the general death-rate, and sewerage brings safety as well as comfort to com- munities, among other ways by ehminating privy vaults and the likelihood of disease transmission by flies, etc.

As to financial considerations, it is difficult to present the full significance of this feature to the general reader without statistics which would be out of place in a short general article like this. Usually water-purification projects proper cost about $3 to $5 per inhabitant served; but pumping, force-mains, conduits, reservoirs, and other associated appurtenances sometimes increase the invest- ment to $15 to $25 per capita. Sewage purification frequently is more expensive than water purification. Upon capitahzing the operating expenses, it is found that modern sanitary works, while involving only small costs for the individual, reach sums of millions and tens of miUions of dollars for our large cities. The solution of these problems has brought many added duties to sanitary author- ities, to professional men who assume the responsibihty for con-

Experimental Methods in Water- and Sewage-Works 23

structing and operating the required works, and to the educational institutions which give technical training to young men desirous of entering this field of work.

new conditions which have been encountered.

Twenty-five years ago the large cities of America were, of course, provided with public water supplies, and many of them had pro- gressed considerably in adopting sewerage systems, although these latter now appear to have been more or less crude. Sewage-purifica- tion plants and water-purification plants, with perhaps half a dozen small and scattered exceptions, were practically unknown. A large proportion of the water supplies, especially in the Central West, were seriously objectionable in the excessive quantities of mud which they carried, and so different in their nature from the comparatively simple filtration projects of Europe that engineers naturally hesitated for financial reasons to attempt their construction, to say nothing of the question whether they would be able to give satisfactory service. The bacterial and hygienic aspects of these problems, which are now recognized to be of so much importance, were almost unknown. In fact, the germ theory of disease had not risen to general acceptance.

The medical man interested in pubHc health knew in a general theoretical way what he wanted, but he was ordinarily unable to state his requirements in a manner understood by the engineer or by the taxpayer. Engineers were able to build any reasonable works, but were unable to learn in terms of the constructor what was required with sufficient definiteness to allow them to make even preliminary sketches and estimates of cost. The chemist and bacteriologist occupying an intermediate position produced with ill-suited methods analytical data full of mystery for everybody.

Misunderstanding continued until men interested in various lines of applied sanitary science co-operated in a manner to make themselves mutually understood. The successful movement to this end, at least in the United States, had its inception chiefly in Massachusetts some 20 years ago. It has resulted during the past dozen years in some notably well-balanced designs for American water- and sewage- works, which have demonstrated their sound

24 * George W. Fuller

merit in practice. In this article it is the endeavor to outHne the development of this aspect of experimental methods.

experimental methods in MASSACHUSETTS.

The experimental methods which have been put in practice in America so much in recent years may be defined as the bringing- together of rehable preliminary data from the engineering, chemical, bacterial, and hygienic standpoints, in order that efficient sanitary works may be built for a wide range of local conditions within the limits of reasonable cost ; and if data are inadequate for fair assump- tions, then the procuring of the needed data by practical tests on a small scale.

It is the Massachusetts State Board of Health to which credit is principally due for developing this method to serve as a guide for such works. In 1886, when the present board was organized, one of its first steps was to establish the Lawrence Experiment Station, whereby data were to be secured to show the best means available under various local conditions for purifying water and sewage. The legislature enacted that this board should serve as a sanitary tribunal, before which the local authorities should place their projects for water- and sewage-works, and whose approval was requisite before the state authorities granted the local author- ities permission to issue bonds for their construction. This experi- ment station has been in continuous service since the autumn of 1887, and has attained a high reputation among various workers in the field of sanitary science throughout the world, in addition to fulfilling its main purpose of aiding the citizens of Massachusetts in economically improving their public works, whereby to a material degree the health and comfort of the people of the state have been enhanced. These results are so well known that it is needless here to go into detail.

The classical investigations at Lawrence, as set forth in the annual reports for the past 15 years, have undoubtedly done more than any other series of investigations in the world to place the science of purifying water and sewage on a sound practical basis. It is true that earlier workers abroad had previously taken important steps along some of these lines, and that sand filtration of water had for

rROPERH UBRART

II, C. State C«U«f«

Experimental Methods in Water- and Sewage-Works 25

many years been in use. But they did not secure comparable paral- lel data from the engineering, chemical, and bacterial standpoints with anything like the completeness obtained at Lawrence, whereby the laws governing successful practice could be broadly stated for a wide range of conditions.

Not only has the Massachusetts State Board of Health availed itself of a testing department, but with other departments it has placed itself in a position to utilize such data advantageously. This has been done by an analytical department procuring data at fre- quent intervals as to the character of various water supplies, rivers, effluents, etc., and, more especially, by a well-trained engineering corps which applies the various data to the needs of each problem coming to the attention of the board.

That the Massachusetts State Board of Health handles well the work coming within its jurisdiction is conceded by all in a posi- tion to know of it intimately. It is true that the board is criticised for not devoting itself more enthusiastically to studies of methods finding favor elsewhere, but this criticism has little to support it. The board properly confines itself to the solution of problems within the state, and of course does not consider it necessary to do more than keep generally familiar with other methods, no matter how suitable they may be for work elsewhere.

EXPERIMENTAL METHODS ELSEWHERE IN AMERICA.

In water purification the Massachusetts problems arc, generally speaking, much easier and simpler than those of the Central West and South, where enormous quantities of silt and clay complicate the necessary works for treating the water, and add materially to the cost as regards both construction and operation. In a manner similar to the procedure at Lawrence, these problems were worked out at LtDuisyille,^ Pittsburg, Cincinnati, and New Orleans. At numerous other places the experimental method has been used in adapting more strictly the design of works to local conditions, especially in the preliminary treatment of turbid waters (Phila- V delphia and Harrisburg), the removal of color from surface waters (Providence), of iron from ground waters (West Superior), the softening of hard waters (Columbus), and the removal of tastes

X

\

26

George W. Fuller

\/

and odors (Reading and Springfield). These problems have all been carefully studied in small test devices for securing data neces- sary for advantageous design and operation.

Sewage purification has also been studied under various local conditions at several places, especially at Worcester, Mass.; Paw- tucket, R. I.; Berlin, Ont.; the Institute of Technology, Boston; Columbus, Ohio; and Waterbury, Conn. In most cases the sewage studies have arisen because of inability or great expense in applying the well-known Massachusetts method of intermittent filtration through sand, or because of peculiarities of the local sewage.

A partial hst of the more prominent investigations as to purifying water and sewage, with dates and approximate costs, is as follows:

List of Special Investigations on Water and Sewage Purification.

Place

Lawrence, Mass

Pro\ndence, R.I

Louiswlle, Ky

Reading, Pa

Pittsburgh, Pa

Cincinnati, Ohio

West Superior, Wis

Washington, D. C

Richmond, Va

New Orleans, La \.

Worcester, Mass. . . .'l.V^O-.'.

Philadelphia, Pa

Springfield, Mass

Harrisburg, Pa

Massachusetts Institute of Technology, Boston.

Columbus, Ohio

Waterburv, Conn

Total

Date

1887 to date 1893-94 1895-97

1897 1897-98 I 898-99 1898-99

1899-1900

1900

1900-1

1900 to Mate 1900-5 1901-3 1903-4

1903 to date 1904-S

1905 to date

Work

Water and sewage Water

Sewage Water

Sewage

Sewage and water

Sewage

Approximate Cost

$175,000 5.000

47-395 I,SOO

36,286

41,588

2,000

8,000

2,000

23,606

37,000

172,000

18,000

25,000

20,000

44,004

10,000

668,379

It is not pretended that the above list is complete. In fact, there are other tests which, while small and of short duration, have had much to do with professional opinion. Perhaps the most important were demonstrations at Louisville and St. Louis, many years ago, that plain sand filtration was incapable of treating the muddy Ohio and Mississippi River waters after plain sedimenta- tion in large basins.

The benefit derived from the experience of the owners of propri- etary devices cannot be overlooked — especially in regard to various appliances of mechanical filters which occasioned the expenditure of much money before being brought to their present state of devel- opment. At Louisville alone the five competing filter companies

Experimental Methods in Water- and Sewage-Works 27

spent more than $50,000. More recently their devices, when tested, have been purchased at the beginning.

While this paper is devoted essentially to methods of purifying water and sewage by works partly or wholly of artificial construc- tion, mention should be made of the important advances in the allied lield of water supply from storage reservoirs, and the disposal of sewage by dilution. Among the more prominent investigations of this kind should be stated those field surveys and laboratory studies made in connection with the Chicago Drainage Canal, the additional water supply of New York City, the improvement of the Charles, Mystic, and Neponset Rivers in Massachusetts, and foreshore pollution along the Massachusetts coast. Several hundred thousands of dollars have been expended on these investigations.

object and advantages of experimental methods.

The purposes of applying experimental methods to problems of water and sewage purification are chiefly threefold, as follows:

1. To provide data for the official and technical authorities, to enable them to adapt new works most advantageously to the local conditions, and to indicate dimensions and other physical conditions permitting contract plans to be prepared and the cost of construction to be approximately estimated.

2. To educate the non-technical public, who as citizens and taxpayers are interested in public works.

3. To provide data so that the officials can operate effectively the works after they are completed, and forecast the approximate cost of operation.

Technical data. — In regard to the first object accompHshed, that of enabling city officials and their technical advisers to design economically works of a suitable character, it goes without saying that this has been of the greatest importance, and is a strong factor in explaining the rapid strides in successful sanitary works accom- plished during the past few years. It has frequently been the advice of technical men, in dealing with problems which differ from those successfully solved elsewhere, to make tests for a year or so at a cost approximating the interest for one year for the works contem- plated. In this way the cost of errors and unbalanced designs

28 George W. Fuller

has been largely minimized, and the efficiency has been increased. In the field of water and sewage purification the information and experience now available are sufficient in a majority of cases to enable these problems to be advantageously handled by experienced workers along these lines. There are some problems, however, that still can be advantageously treated by the experimental method. They refer especially to sewage problems in which trade wastes are involved, and to water problems where the composition of the water is quite unusual in some particulars on frequent occasions. From the technical standpoint, however, the field of water and sewage purification, broadly speaking, has passed beyond the experi- mental stage, and the advances, both as to efficiency and economy, are largely to be gained, not from experimental plants, but by the careful and more systematic operation of works in practice. Such studies will, of course, lead to improvements which can be taken advantage of in the construction of new works, and will gradually bring to a higher plane of excellence the art of water and sewage purification on its present scientific basis.

Educational aspect. — It is frequently said that communities progress in proportion to the advance in knowledge of the average citizen, or to the mean knowledge of the community as a whole. There is a good deal in this, and it brings forcibly to mind the neces- sity of educating the pubhc as to what improved sanitary conditions really mean, and of letting them ascertain for themselves what can be accompHshed along these hnes in the field of applied science. Non-technical people have a natural aversion to the word "experi- ment," notwithstanding the aid derived from devices which not improperly may be termed experimental. While the term "experi- ment station" from its use at Lawrence and a few other places seems to have a firm footing in some localities, it is gradually giving place to the term "testing station." This is a much preferable expression in many ways, as it disarms the criticism of many who seem to think that these investigations are conducted in a "hit or miss" manner, much after the fashion of the early inventors. This is not so, as experimental methods, as now ordinarily apphed to water- and sewage-works, are aimed at testing procedures found successful elsewhere, but which may require adaptation to local

Experimental Methods in Water- and Sewage-Works 29

conditions in regard to some details. Their magnitude, while rela- tively small for reasons of economy, is still much greater than that which seems to be taken for granted by numerous citizens, who associate the word "experiment" with a test tube, or with a mechani- cal device which is so imperfect that no one dares to build it on a large scale without further experiments. The methods of puri- fying water and sewage have now advanced to a degree where the phrase "testing station" in new projects will unquestionably displace "experiment station;" and the testing of these processes where unusual conditions are expected will assume a dignity comparable with that of the regular departments which systematically test cement, steel, and other materials used for building purposes. In fact, it is interesting to note that the laboratories at many testing stations have been utilized regularly for testing construction materials.

Where water- and sewage-purification projects involve hundreds of thousands of dollars or more for construction costs, the so-called experimental methods, as applied in accordance with the foregoing statements, have given wonderful courage in many places to offi- cials who otherwise would very naturally have been in a hesitating frame of mind, and inclined more to listen to the "doubting Thom- ases" who in all communities, for selfish or other reasons, appear as opponents and obstructionists to modern sanitary works. Even if the technical advisers of the projects were not assisted by such data, it is quite likely that the testing station for many projects would indirectly in this way do far more good than the cost involved, in saving lives and in hastening the day when communities will meet their problems in accordance with the best information available.

In speaking of the educational benefit derived from applying experimental methods to water- and sewage-works, the technical men, especially those in charge of the tests, have an important duty to perform in teaching non-technical officials, and various citizens who are interested in the work, the fundamental principles of the process involved, and in assisting them in ascertaining what prac- tical works would mean, both hygienically and financially. Along this general line the Institute of Technology has played an impor- tant role, largely through having had for many years on its teaching stafi" a man who to an extraordinary degree possesses the faculty

30 , George W. Fuller

of getting fundamental truths of sanitary science before his hearers in such an attractive manner that they never forget them. It is the behef of the writer that the vi^ork accomplished by Professor Sedgwick along this line is unequaled by that of any other man in this country, either in educational or other lines, and that this fact in a few years will be far more widely realized than at present, when his younger pupils throughout the country reach an age where their work will be felt in the communities in which they live. This influence is already to be found in many unexpected places, and forms a wonderful tribute to the success accomplished by Professor Sedgwick in one of his many lines of usefulness.

Operation oj works. — After water-and sewage-purification works are constructed, it is imperative that they shall be operated in an intelligent and efficient manner. The benefit of this has long been demonstrated in Europe, and the absence of such supervision in many places in America shows the folly of careless and indifferent management. No matter how well water- and sewage-purification works may be designed and built, there is no engineer who can give assurance that the results accomplished will be satisfactory unless the works are well managed. Not only must the works produce a result which is satisfactory from a scientific standpoint, but their behavior should be put before the citizens in a way that will inspire confidence. When fair-minded citizens as a mass continue to lack confidence in works of this type, the latter cannot be called an unqualified success, no matter how fully scientific facts may show their adequacy.

The Massachusetts Institute of Technology instituted the plan of especially training young men along technical lines, so that they might become competent to serve as superintendents for water and sewage-purification works. In this pioneer work they are entitled to great credit, and their example is already being followed by similar institutions elsewhere. This is an important field of technical education, as a majority of such technically educated men in the future will be connected with the management, rather than with the construction, of works of this type.

In passing, it may not be amiss to say that the technical managers of works of the type under consideration must have other quali-

Experimental Methods in Water- and Sewage-Works 31

fications than those of a scientific nature. They must be able to maintain amicable relations with executive superiors, to manage laborers, to keep records in a manner fairly comparable with the high degree to which the art of bookkeeping in large business houses has advanced, to prepare reports containing essential features in explicit but terse terms, and to make plain to non-technical men in both public and private capacity the more essential features of their own position and of the data by which their efforts show what is being accomplished. This type of specialists will naturally develop in efficiency as their responsibihties increase; but there is still much work for the technical schools to do in preparing young men more adequately for these duties.

Tentative installations. — As distinguished from the testing stations built solely for the purpose of tests, there is, of course, one other method of a somewhat experimental nature by which local data are used in determining whether large works are most advanta- geously constructed. I refer to the plan of constructing works gradually, or tentatively, and of using data from the operation of the first portion of the installation to serve as a guide in arranging the details of the portions subsequently to be built, and also in deciding upon the magnitude of the works sufficient for a given capacity or to serve for a given term of years. This is the style of works, from the experimental point of view, which frequently obtains in Europe, and which will obtain in some places in this country. As yet there has not been a wide apphcation in America of such data obtained on a large practical scale, although, of course, they are availed of more or less in all works where extensions are required. This condition has been reached at several sewage- works in New England, and the results of experiences in the field have been summarized by the Massachusetts State Board of Health. It is gratifying to state that practical results are in general con- formity with the principles of water and sewage purification as developed by tests on a small scale.

experimental methods in EUROPE,

In Europe the water-purification problems do not cover nearly so wide or difficult a range of natural conditions as those met in America. Filtration has in recent years not received as much attention experi-

32 George W. Fuller

mentally as has been the case with sewage-works. In earher years however, experimental methods had much to do with the develop- ment of water filters abroad. It is not to be forgotten, furthermore, that in Germany much good work during the past dozen years has been done in developing the most practical methods for removing iron from ground waters. At present the most interesting feature of water-purification developments in Europe refers to the prelimi- nary treatment for some of the river waters which are fairly turbid during freshets, and to efforts to sterilize water economically by ozone. The most notable instance of the former is at Suresnes, near Paris, where the Seine water below the metropolis is subjected to filtration six times, the first filters being of coarse gravel to effect clarification.

In England, which is the home of modern sanitary engineering, sewage-purification works have received more attention than in any other country. The density of population in England and the relatively small size of its rivers have, of course, forced this con- dition at an earlier date than is generally true of other countries. While for some years the English have not contributed much on the subject of water filtration, their experience in the field of sewage purification far exceeds that of any other country. Experimental methods in one form or another have played an important part for half a century, beginning with efforts to utihze the manurial value of sewage. This is largely owing to the differences in various local conditions, especially topography, geology, and the compo- sition of the sewage as influenced by trade wastes. Not only have the English conducted test filters and other processes of purification on a small scale, but they have also gathered many data of great value by the operation of their works in practice along Hnes which enable current experiences to be utilized in developing future works.

These data have been so universally obtained in conjunction with the operation of existing works in practice that it is very difficult to ascertain even roughly what their cost has been. The staff regu- larly engaged in operating the main works has secured the technical data, so that the expense has been confined to building the test devices, relatively small in size, and to a Httle extra labor for opera- tion. The large mass of valuable testimony published in numerous

Experimental Methods in Water- and Sewage-Works 33

municij)al reports and by the Royal Commission on Sewage Dis- posal shows what a fund of knowledge has been accumulated at London, Salford, Sutton, Exeter, Burnley, Accrington, Hudders- field, Leicester, Birmingham, Bradford, Devizes, Hanley, and other cities, and which for most places has been obtained with almost no special fund devoted to testing purposes, comparatively speaking.

At Leeds the unusually thorough sewage tests made during the past eight years received appropriations of about $150,000, some two-thirds of which has been actually devoted to that purpose. Manchester has also expended quite large sums for experimental purposes, although, for the reasons above stated, the expenditures were by no means commensurate with the information obtained. The Royal Commission on Sewage Disposal in England is under- stood to have an appropriation of about $55,000 for the expenses of its own stafif and the traveling expenses of the numerous witnesses who have appeared before it. There are also special river boards and county councils, with excellent technical staffs, which gather many valuable data.

In France sewage purification has been the subject of experi- mental study, beginning with the labors of Mille in 1868 at Gen- nevilliers. These tests resulted in the establishment of the present sewage farms of Paris. Within the past few years the biological methods of purification have received attention both from the city of Paris and from the Department of Agriculture. The latter has a general supervising control over water and sewage matters outside of Paris, and is devoting an appropriation of about $60,000 to such investigations. Thus far these studies have been made by Professor Calmette at Lille, as set forth in his interesting progress report of last autumn.

In Belgium the government is paying particular attention experi- mentally to the treatment of trade wastes at a special station devoted to that purpose at Verviers.

The government of Holland established, in 1904, a sewage-testing station at Tilburg, the cost of which to date is approximately $15,000. No reports have yet been pubhshed. Several ozone plants have been tested in Holland, and the city of Rotterdam is now arranging to test a mechanical filter on the local river-water supply.

34 George W. Fuller

In Germany numerous experiments have been made upon the sedimentation of sewage for purposes of clarification, and the so- called biological methods have been studied for some years, begin- ning in 1895, when a testing station was estabhshed by Professor Dunbar at Hamburg, which station is still in operation. In 1901 the Prussian government estabhshed a permanent organization for testing water- and sewage-purification methods. This "insti- tute" has gathered together and pubhshed the more important data as to experiences in other countries, has conducted several important sewage-testing stations in the suburbs of Berlin, and has collated many useful data as to the sanitary works of the cities of Prussia and neighboring territory. This department has an annual appropriation of about $30,000 for testing, inspecting, analytical, and clerical purposes. The sum devoted to testing purposes varies, but is materially supplemented by the arrangement of conducting investigations for various local authorities, the expense for which is borne in part by the community benefited. The department also established the custom of officially examining proprietary devices, largely at the expense of the owners in cases where the devices seem to possess sufficient merit. In this way a mechanical filter of the Jewell type was recently tested at the Miiggelsee plant of the Berlin water-works. The same filter is now being tested on the colored water supply of Konigsberg.

The relative amounts of suspended matters deposited from sew- age at different velocities have been studied carefully under vary- ing local conditions at Frankfurt, Cassel, Hannover, and Cologne, as shown by the data published in municipal reports and the technical press. In these cities, as in England, it is difficult to ascertain the cost of the tests, because so much of the work was done by the regular staff of the technical authorities of the cities. The scope of the tests has probably been greatest at Frankfurt, including means for most easily removing sludge, its partial drying by centrifugali- zation, and its ultimate disposal by incineration after mixing with the city refuse. About $60,000 has been spent at Frankfurt on these and other sewage tests, including filtration, within the past dozen years or more.

Professor Dunbar's activities in the field of sewage purification

Experimental Methods in Water- and Sewage-Works 35

have by no means been confined to Hamburg. His publications show that he has advised the authorities at Miihlhausen, Stuttgart, Beuthen, Unna, Leipzig, and other places. In nearly every instance he has taken advantage of experimental data to ascertain local conditions. Leipzig and Chemnitz in Saxony are now conducting sewage tests, the appropriations for which are about $17,000 in each case, with the engineering and analytical data secured by men regularly employed by the city.

This brief record of experimental methods as applied to water and sewage purification can hardly be brought to a close without reference to trade wastes. This feature in aggravated cases com- plicates the design of sewage-works and adds materially to the costs of operation. Various industries require special consideration, as shown by the efforts of the river boards to minimize the effect of trade wastes in the streams of Lancashire and Yorkshire, England. The removal of fats has perhaps received the most attention along this hne — especially in Berlin, Cassel, and Chemnitz in Germany, Verviers in Belgium, Roubaix and Grimonpont in France, and Bradford, Manchester, and Oldham in England. Numerous mill- owners also recover grease from their waste water. The extent of some of these investigations is indicated by the fact that at Cassel a private company is said to have spent considerably more than $100,000 in unsuccessfully endeavoring to fulfil a contract for extract- ing fats from the city sewage. The only large place where the entire city sewage is regularly treated for grease extraction is at Bradford, England.

THE FUTILITY OF A SANITARY WATER ANALYSIS AS A TEST OF POTABILITY.*

Marshall O. Leighton.

Whosoever expresses doubts concerning generally accepted ideas must be prepared to see his statements misinterpreted and their application carried far beyond the point at which they were aimed, even to the absurd and grotesque. He must not expect that his observations and deductions will be confined to the Hmits prescribed, even though he resorts to every safeguard that his mother- tongue affords. More attention is paid to the devious paths along which his statements may lead by implication than to the single trail that he has defined by precise guide-posts. Finally, such a person must sustain confrontation by that splendid, indispensable, and all-saving power known as conservatism. Therefore the author of this paper hoists a flag of truce while he makes his preHminary declaration, in the hope that the highest possible proportion of those interested may not mistake his line of march.

1. All contentions concerning the futility of sanitary analyses are applied strictly to waters. Sewages, fresh and stale, and sewage effluents are expressly ehminated from consideration, except in certain cases where they will be taken to illustrate the fact that they may occasionally be accepted as unpolluted water, according to standard methods of interpretation.

2. // is not contended that all sanitary water analyses are futile irrespective of the conditions under which they are made and inter- preted. In consistent studies of nitrogen, as such, and the changes which take place in its form, such analyses are important.

3. // is admitted that in certain limited areas of the United States sanitary water analyses afford information by which animal pollu- tion may occasionally be detected.

4. The facts comprised in the foregoing admission have been a stumbling-block to chemists working with waters outside of those limited areas.

* Received for publication, March 30, 1906.

36

Futility of a Sanitary Water Analysis 37

5. // is contended that the sanitary analysis offers nothing by which one may positively distinguish between a dangerous and a whole-^ some water.

6. The composition ratios that many good men cherish may be applied indiscriminately to wholesome waters and dilute sewages.

7. The conventional method of seeking for evidences of pollution by sanitary analyses, or of accepting or rejecting a water upon such evidence, is in its broad and essential features quite misleading, too frequently dishonest, and in some cases absurd.

8. The dangerous pollution of surface waters can be discovered more readily, and at far less risk and expense, than by sanitary an- alysis.

9. The term "sanitary analysis" as used in this discussion does not include tests for specific organisms.

Standards of interpretation by which a water may be designated as "good" are faithfully met by many waters undeniably bad; con- versely the characteristics of a water interpreted as "bad" are pre- sented by many the wholesomeness of which cannot be questioned.

There is in the presidential address of Professor Leonard P. Kinnicutt, delivered before Section C of the American Association for the Advancement of Science, at New Orleans, La., in December, 1905, a concrete statement of intrepretation standards. This state- ment will be used as a basis for the comparisons which follow in this discussion. Such a selection is made, decidedly not for the purpose of controverting the statements or discrediting the position of this distinguished authority, but rather because it is the most admirable resume that has recently emanated from a highly respected and competent source. The following statements are quoted:

(A) In fresh sewage the amount of nitrogen as free ammonia is from three to four times that of the nitrogen in the albuminoid ammonia, and in sewage efHuents from 20 to 30 times, while in peaty water, or water containing an infusion of leaves' the nitrogen in the albuminoid ammonia is from lo to 20 times the nitrogen in free ammonia. Hence, when a surface water, not including rain or snow water, gives a greater amount of nitrogen as free ammonia than it does as albuminoid am- monia, the indications arc that the water has certainly been polluted by sewage, and that the source of the organic matter is of animal origin. With a large amount o^ nitrogen as albuminoid ammonia (over 0.25 milhgram per liter) a ratio of nitrogen of the free ammonia to the nitrogen of the albuminoid ammonia of less than i to 5 is suspicious.

////

3^

Marshall O. Leighton

(B) Consequently, while a low ratio as i to 5 between the nitrogen of the free ammonia and the nitrogen of the albuminoid ammonia indicates pollution, the reverse cannot be said to be a strong indication that the water is a normal water.

(C) A colorless water containing that amount of nitrogenous matter represented by 0.25 milligram of nitrogen as albuminoid ammonia per liter is looked upon with suspicion.

(D) Free ammonia always indicates organic matter in the process of decompo- sition. In unpolluted surface waters it is rarely high, being removed almost aS fast as formed by vegetable and animal organisms in the water, and an amount of nitrogen as free ammonia above o . 05 milligram per liter is unusual, and, if it does occur, the water cannot be considered as an unpolluted water unless that fact is clearly established by other data.

Attention is then called to seasonal variations and the increase in free ammonia during the autumn in northern countries.

(E) Concerning nitrogen as nitrites:

More than 0.002 milligram per Hter is an unfavorable indication.

(F) Concerning nitrogen as nitrates:

It is never present in any large amount, seldom exceeding o . i milligram per liter. Higher amounts than this, being unusual, must be looked upon with suspicion.

Professor Kinnicutt then explains that the above interpretations refer to reservoir, pond, and lake waters, but that in river waters

high nitrogen as free ammonia, as albuminoid ammonia, and as nitrites charac- teristic of recent pollution in ponds and reservoirs may be due to the decomposition of algae life, which was stimulated by the entrance of sewage in the upper stretches of the river.

Accepting the above as an authoritative basis of interpretation — and it is the one which closely corresponds to that which the writer has found in very general use — let us interpret a few analyses. Ref- erence will be made by letter to the foregoing quotations, so that the basis of each interpretation may be clearly defined.

SERIES "A." Parts per Million.

Date

Tur- bidity

Color

Odor

Nitrogen as —

No.

Albuminoid Ammonia

Free Ammonia

Nitrites

Nitrates

Chlorine

1 2 3

July II, 1900 July 20, 1899 Sept. 14, 1900

0 Cons.

SI.

0

1-5 2. 1

0 im im

0.026

0.330 0. 114

0.028 0. 246 0. 164

0 0 0

1.2 1 . 2 I. 2

There are presented in the above series three analyses of pond waters. All are practically colorless, and Nos. 2 and 3 have a

Futility of a Sanitary Water Analysis

39

slightly moldy odor. (A) In Nos. i and 3 the nitrogen as free ammonia is greater than that as albuminoid ammonia. (D) Nos. 2 and 3 contain very much greater amounts of free ammonia than 0.05 part per million. (A, second part) No. 3 contains over 0.25 part of albuminoid ammonia per million, and the free- albuminoid ratio is i to 1.3. No nitrites or nitrates appear in any of the samples. According to the above standards of interpreta- tion, all three of these waters contain recent organic pollution of a very unstable nature or, in other words, sewage pollution; the nitrification is proceeding very rapidly, and the assimilation of nitrites and nitrates is accomplished as rapidly as they are formed, by an abundance of organisms. In point of fact, these are normal waters from two storage ponds in Pennsylvania, in the drainage areas of which there are no habitations. It would be difficult to specify con- ditions that would more closely approach the ideal for upland con- served supply than existed at these two places at the time these sam- ples were taken. No. i is from Pine Run Reservoir, and Nos. 2 and 3 from Mill Creek Reservoir, both in Luzerne County, Penn- sylvania.

SERIES "B." Parts per Million.

No.

Tur- bidity

Color

Odor

Nitrogen as —

Chlorine

Total Residue

Imss on

Albuminoid Ammonia

Free Ammonia

Nitrites

Nitrates

Ignition

I 2 3 4 5

DLst. It

It

((

(t

13

IS 8 8

13

2a la 2a la la

0.120 0. 204 0. 106 0. 142 0.130

0.006 0.016 0.002 0.026 0.012

0.000 0.000 0.000 0.000 0.000

0.170 0. 100 0. 160 0. 160 0. 160

0.9 0.9 0.9 0.9 0.9

69 S 72.0 66. s 66.5 69.0

It). SO 20.00 16.00 13 so 20.00

Series "B" contains analyses of samples taken from a large lake in September, 1904. Each sample was distinctly turbid, of low color, and revealed an aromatic odor. (C) The nitrogen as albuminoid ammonia is in every case less than 0.25 miUigram per liter. (D) The nitrogen as free ammonia is in all cases far less than 0.05 milli- gram per liter. (A) The free-albuminoid ratio varies from 1-5.3 up to 1-5.5. (E) There are no nitrites. (F) The nitrates run somewhat higher than the standard set.

Several of the above samples have all the characteristics of an

40

Marshall O. Leighton

infusion of leaves (A) except color. There is nothing in the analyses presented, except the unimportant excess of nitrates, that does not surpass on the acceptable side the strictest of the interpretation standards above quoted. The samples were collected almost simul- taneously on September 22, 1904, from Lake Champlain, within the inclosed area lying between the docks at Burlington, Vt., and the harbor breakwater. Sample No. 2 was taken about 1,000 feet away from the outlet of the main trunk sewer of the city. The remain- der were taken at points less than 500 feet away fom said outlet. No. 5 being collected about 20 feet from the sewer's mouth. Ten years previous to the collection of these samples the city of Burlington was obliged to abandon a water intake situated at a much more favorable point than those at which any of the above samples were taken. The reason for the change was the high rate of intes- tinal disease morbidity in the city, which was markedly decreased afterw^ard. What, then, shall we say of sanitary analysis as an index of pollution at Burlington ? Bacteriological examination reveal- ed the abundant presence of B. coli in all the samples reported in Series "B," and the discharge of sewage into the lake was a matter of casual observation. Therefore no one was deceived by the nitro- gen determinations. One may very properly question whether sanitary analyses may not be, under less fortunate circumstances, an actual danger to public health. Let us now examine series "C."

SERIES "C." Parts per Million.

No.

Nitrogen as —

CUorine

Total Residue

Loss on Ignition

Albuminoid Ammonia

Free

Ammonia

Nitrites

Nitrates

I

2 3 4 5

0.17 0. 10 0.07 0.18 0.06

0.01 0.01 0. 12 0.09 tr.

0 0 0 0 0

1 .00 1. 00 I 25 2.25 3 00

2-5

50 4.0

50 50

83

78

219

73 32

25

37 42 24 15

Records of color and odor are unfortunately absent, but from independent sources comes the assurance that none of the samples were highly colored, and the predominating odor is faintly earthy. According to the standards of interpretation, Nos. i and 2 were in good

Futility of a Sanitary Water Analysis

41

condition at the time of analysis (C and D). The ammonias arc low, and (A) the free-albuminoid ratio is excellent. (F) It would appear from the large amount of nitrates and the high chlorine that the water has at some time been polluted, but has become well purified. Nos. 3 and 4 are waters that have been in bad "company" (F), and the free-albuminoid ratios, especially that of No. 3, show recent pollution (A). No. 5 (see high nitrates and chlo- rine) appears to be an excellent example of a water purified by run- ning in a stream-bed through a long stretch of unoccupied country below some initial seat of infection. Note the high nitrates. The truth is that all these waters were taken from unpolluted streams in the mountain districts of the Potomac drainage area in Virginia and West Virginia.

SERIES "D." Parts per Million.

				Nitrogen as —		u ■0	Hardness	X
Date	■B a	rt					0
	13	"o U	1_ 0	II 1^	a 0	C/1 2	2;	3 0	e2	>, •a	J3	3 2
Aug. 22, 1903. .	1	4	0	0.056	0.050	0.	0.025	1.73	68	46	8.3		1	I
Sept. 28, 1903. .	40	l.S	0	0.341	0.055	0.	0. 125	3.64	121	52	23.0		7	2
Oct. 24. 1903. .	5	26	0	0. 140	0.04	0.	0. 125	S.90	130	32	330		35	3
Nov. 17, 1903. .	60	."il	M	0. 226	0.084	O.OOI	0.144	S-io	144	28	330		3	4
Dec. 8, 1903. .	I	9	0	0.054	0.034	0.	0.025	2.80	77	29	3SO	1:15	S

Analysis No. i, in Series "D," indicates a practically colorless and odorless water, with nitrogen in all four forms low in amount. The chlorine, too, is low and practically, the only suspicious feature about the statement is the free-albuminoid ratio (A).

No. 2 is a turbid water of moderate color. The amount of nitro- gen as albuminoid ammonia is high, but the free-albuminoid ratio ("A," last part) is 1:7. (F) Nitrogen as nitrates is high. It is not a very bad water according to the interpretation, yet there are evi- dences that some swamps are tributary to the point at which it was taken.

No. 3 looks suspicious because of the free-albuminoid ratio (A), the moderately high free ammonia (D) and nitrates (F), and the high chlorine.

42

Marshall O. Leighton

No. 4 is a turbid, highly colored water, with a moldy odor, a bad free-albuminoid ratio (A), an appearance of nitrites, and high nitrates (F) and chlorine. It is a thoroughly " suspicious "-looking water.

No. 5 has practically the same characteristics as No. i.

The object of introducing these tables is not so much to show the misleading character of the data as to call attention to their variations. Here are five analyses of a normal water, taken at the same point from a small stream draining an uninhabited wilderness. Yet only two of them possess a resemblance of uniformity, and the free-albuminoid ratios vary from those of a dilute sewage to those of a potable water. The variations in chlorine, too, are interesting, and they lead one to speculate upon the actual normal chlorine value for this region. The samples were taken from the head waters of Green River in Casey County, Kentucky.

					SERIES " E	M
					Parts per Million.
					Nitrogen as —		3	Hardness	T3 1

				■a rt	C3				-a			E Is
Date				'§'3 .S ° 6 2	•a 0 E			J		•a		«
	3	S	0 ■0		^A	'u.		0 2	3 ,0			v.^,	E 3
	H	0	0	<	£	z	2	u	h	<	l^	fa	Z
Sept. 2o, 1903	20	20	0	0.272	0.130	0.009	0.22s	S.6	130	6s	24	1:2	I
	20	17	0	0.302	0.130	0.002	0.187	5-5	128	60	14	1:2.3	2

Series "E" presents several points of interest. Both waters are of moderate color and turbidity, and have no odor. According to the standards of interpretation. No. i is a recently polluted water. It contains a large amount of nitrogen as albuminoid ammonia, (A) and the free albuminoid ratio is i to 2. Free ammonia is very high (D). Nitrites and nitrates (E and F) are both high. On the whole, the water may be said to be both recently and remotely pol- luted. No. 2, although somewhat similar, is superior in some respects. The free-albuminoid ratio is i to 2.3. Free ammonia is the same, while nitrites, nitrates, and chlorine are lower, though the last is not significantly different. The fact that in No. 2 the albuminoid ammonia is higher than No. i is responsible for the better ratio in the former.

One might readily infer that both samples were taken from the

Futility of a Sanitary Water Analysis

43

same stream at the same place. "Bad" water No. i was taken from Kentucky River above the city of Frankfort at the water-works intake. "Bad" water No. 2 is from Kentucky River below the Frankfort sewers. Note that the dates of sampHng are the same. The important feature of Series "E" is that here is a water, or a dilute sewage, taken below the sewers of a city of 20,000 inhabitants showing practically the same, if not a better condition, according to the interpretation standards, than another sample taken from the stream above the sewers.

Cases similar to the above are very common. Two good examples are presented in Series "F."

SERIES " F " Parts per Million.

Date

Nitrogen as —

Hardness

!2

3

.3

■3

1

0

-0 ■3 -a

1

5^

0

E

a* G

1

•n

0

IS 0

-a 1

•a

1 IS 1=

e

•i

B

a

Mississippi River at Brainerd, Minn.

Nov. 3, 1904

(( (( ii

Feb. 28, 1905

14	112	2V	0.382	0.022	0	0.03	I.O	ISO	100	4	1:17
IS	112	2V	0.382	0.040	0	0.03	I.O	142	100	4	1:9.5
10	40	IV	0.256	0.051	0.002	0.04	1-4	i8S	161		i:S
10	40	iv + iM	0.300	0.040	tr.	0.04	1.8	188	162		1:75

Above town Below •' Above " Below "

St. Louis River at Cloquet, Minn.

Oct. 31, 1904 Feb. 25, 1905

17	292	2V	0.502	0.036	0	0.04	1.6	121	34	I	1:14
19	112	2V	0.302	0.032	tr.	0.08	1.8	1S6	98		1:9.4
-7	112	2V	0.322	0.152	tr.	0.08	1 .2	I SI	104	4	1:2
-7	112	2V	0.302	0.032	tr.	0.08	1.8	156	98	7	i: 10

Above town Below ' Above ' Below

SERIES "G." Parts per Million.

>•

■| 13

u

3

Nitrogen as —

.3

u

0 u

3

e2

1

B

Is

E

3

0

■0.3

5

0

B -J

2

8

Date

1. .

2. .

3--

4.. $■■

6..

7.-

SI. Sl. SI. SI.

si.

SL

0 0 0

0. 1 0

0. 1 0. 2

0.480 0.37s 0.656

0395 0.53s I . 200 0. 511

0.080 0.013 0. 104 0.026 0.026 0033 0.085

0 0 0 0 0 0 0

0.461

0.37s

0.307 0. 142 0.472 0.349

70 76

59 74 67

S7 68

634 612 818 790 566 850 604

1:6

1:29

1:6

i:iS 1:21 1:36 1:5

Mar. s. 1808 " II, 1898

Apr. 9, 1898

Mar. II, 1898 " 18. I«9» " 25. 1898

Apr. I, 1898

44 Marshall O. Leighton

The waters represented in Series "G" are typical of the western prairie states. In such waters large amounts of organic matter are always present even in the absolutely uninhabited regions. The sam- ples above reported are taken from a stream that drains a large area underlaid with saline deposits, which account for the high chlorine content. Such waters absolutely controvert paragraph (C) of the foregoing interpretations. It will be noted that the free-albuminoid ratio specified in "A" (last part) is satisfied by all of the analyses. There are no nitrites, and only in Nos. i, 3, and 7 does the amount of nitrogen as free ammonia exceed the standard set in " D." Nitro- gen as albuminoid ammonia and nitrates are not high for prairie waters. Although some of the analyses look "good," the samples were all grossly polluted. The first three samples were taken from Kaw River i . 5 miles above Lawrence, Kans., and contain the residue of pollution from the sewers of Topeka. The last four samples ware taken from the same stream, but the point was 300 feet below the outlet sewer of Lawrence. It will be seen that the samples from below town present a better analytical appearance than those from above.

We will now consider ground waters. In the address above quoted there are the following statements:

(A) .... It may be said that the best ground waters should certainly con- tain not over o. oi milligram of nitrogen as free ammonia, or (B) over o . 02 milligram of nitrogen as albuminoid ammonia, (C) no nitrogen as nitrites, (D) not over o . 01 milligram of nitrogen as nitrates in a liter of water, and (F) chlorine not above the normal of the region. When a water contains (F) more than 0.05 milligram of nitrogen as free ammonia, and (G) o . 08 milligram of nitrogen as albuminoid ammonia, or 0.12 milligram of nitrogen as albuminoid ammonia, even if the free ammonia occurs in very small amounts, it is a sign of imperfect filtration or of subsequent pollution, and consequently such water should not be used for household purposes.

It is more difficult to determine the presence of pollution in a ground water by inspection than in a surface water, and in dis- cussing ground-water analyses one is sometimes unable to make a definite statement concerning direct pollution. We can, for exam- ple, in the case of a surface water state that if the water is from an uninhabited region it must be unpolluted with animal waste. On the other hand, there is but one certain method of determining the healthfulness of a ground water. This method has been accepted

Futility of a Sanitary Water Analysis

45

by chemists and sanitarians the world over as being the best test of the wholesomcncss of all waters, whether from the surface or from the ground — namely, the incidence of typhoid fever and other water-borne diseases among the habitual users of such water as a beverage. Although there are on record many hundred analyses of ground waters, which would, by the interpretation standards above set forth, be classified as polluted, but which, judging from the location and all the surroundings, might be regarded as wholesome, never- theless the basis of the statements made in the following paragraphs will rest solely upon the typhoid rate prevailing among the users of the various waters. The first series of ground-water analyses to be discussed are grouped in the following table:

SERIES "H." Parts per Million.

Nitrogen as —

Chlorine

Total Residue

Date

Albuminoid Ammonia

Free Ammonia

Nitrites

Nitrates

Aug. 30, 1905. . . .

Oct. 10, 1900

May 18, 1905

" 18, " ....

May II, 1898

Dec. 8, 1904

0.044 0.082 0.048 0.216 0.058 0. 192

0.136 0.054 0.032 0.032 0.062 0.032

0.008 0.006 0.009 0.009 0.007 0

0.152 9 394 9.000 4. 200 0.600 0.080

1.80

6.40

12.00

63.25

10.00

7.50

341 2 324. 8 427.2 1042.4 432.8 3S30

It will be seen from the above that all the waters analyzed con- tained more nitrogen in all the specified forms than would be allow- able under the standards of interpretation above quoted. The analyses presented represent either the city supply of Rockford, III., or that from private wells which are largely used in that place. They are in all cases ground waters, and are similar in character to waters from various wells in that region. The writer has before him 135 analyses of well waters from Rockford, by far the majority of which present characteristics similar to those presented in Series "H." Rockford has the lowest typhoid fever death-rate of any city in the United States having a population of 30,000 or over. It will be noted in the various pubhshed tabular statements, such as that presented by Mr. George C. Whij)ple in the report of the Commission on Additional Water Supply for the City of New York that Rockford almost invariably stands at the foot of the list, with a death-rate of about 6 per 100,000.

46

Marshall O. Leighton

Let us consider another case:

SERIES "I." Parts per Million.

Organic Nitrogen

Nitrogen as —

Chlorine

Total Residue

Oxygen

Consumed

Albuminoid Ammonia

Free Ammonia

Nitrites

Nitrates

2.75 2.86 2.68

0.280

I. TOO

0.610

0.07 1. 100 0. 142

0.001

O.OOI O.OOI

0.57 0.24

0.37

7.5 12.0

6.3

368 569 400

3-73 301 2.6s

The analyses in the above table represent the city water of Des Moines, Iowa. The first represents the water from the large well; the second, from the small well; while the third is an average of 42 analyses of the supply, all made in the year 1897. Here again comment upon the divergence of these figures with those given in the standards of interpretation is unnecessary. Des Moines, Iowa, is one of the most fortunate cities in the country from the stand- point of typhoid rates.

Series " J " contains respectively the average of analyses made from the Oconee and the Shetucket wells of the Brooklyn water supply. Throughout the entire period between 1897 and 1902 it will be noted that in neither case does the average number of bacteria exceed 50 per c.c, and there were no positive tests for coli during the entire period of investigation. Nevertheless, the nitrogen determinations^ according to the above standards of interpretation, would condemn this water.

SERIES "J." Parts per Million.

				Nitrogen as —			Hardness		u	4j *.»
				•0.2	cfl				3 •a		m		K	l"*^
	>,			■33	"a				C/3	>t	u		Q
	'■3			si	E	£	ffi		a	'c			.2
	J2 5	"o		-1	b	13		0	"3	ES3	a 0	u	er c test in c
	H	0	0	^	£	2	K	U	H	^	:<	(— 1	m	Oh
Oconee WeUs . . .	I	6	0	0.20	2.65	O.OOI	O.OI	4.8	148.7	lOI . 0	50	0.57	33	0
Shetucket Wells .	10	25	0	o.ois	0402	0.012	O.OI	264.2	713-5	80.6	1594	2.02	50	0

Another example is contained in Series "J." The samples were taken from an isolated well at St. Cloud, Minn., and it must be con- fessed that the analyses have an unfavorable appearance, especially

Futility of a Sanitary Water Analysis

47

by reason of the amounts of nitrites. The water, however, is abso- lutely unpolluted. The first sample contained two bacteria, and the second, one bacterium per c.c. Only one species was represented, which was not B. coli.

SERIES "T." Parts per Million.

	Date	>.			Nitrogen as —		3	Hardness
	■0.2 C 0	.2 '5 0			>.
		"2 IS 3	7!	0 •a	11 5	E ''J	10	1	_o	0		11
		H	U	U	Uh	'i^	z	U	H	<	2;
Sept. 4,	1904	0	0	0	0.042	0.014	0.015	7.00	3-0	256	156	17
" 4.	*'	0	0	0	0.036	0.014	0.066	7.00	30	276	IS7	16

For a final example the following analysis is submitted, the expres- sion of results being in terms of parts per million :

Free ammonia 0.012

Albuminoid ammonia 0.012

Nitrites . . 0.00

Nitrates "strong"

The above water is from a well in Rochester, Minn. It will be seen that the ammonias are extremely low in amount, and the nitrites and nitrates practically absent. It is a water which con- forms in all respects to the standards above quoted, yet it contained 1,570 bacteria per c.c, and an abundance of B. coli.

There remain for consideration the artesian or deep-seated rock waters. Examples might be cited in the support of the general contention of this paper, but it will be distinctly preferable merely to refer to a paragraph in the address above quoted, as follows:

Unfortunately, however, in the study of artesian water perplexing chemicaj and bacteriological results are often obtained. In artesian waters so situated that surface pollution seems impossible, amounts of nitrogen as free ammonia, as nitrites, and as nitrates have often been found which, if occurring in ground waters, would cause them to be considered as polluted. The nitrogen of the nitrates in these waters may be due to fossil remains, and the nitrogen as nitrites and as free ammonia to the reduction of the nitrates by chemical action, as contact with iron sulphide, and the occurrence of the nitrogen as free ammonia also sometimes to some salt of ammonia existing in the strata through which the ground water passes. On this account the determination of the nitrogen content does not give as satisfactor}' data from which to draw conclusions as those obtained from the analysis of ground water.

48 Marshall O. Leighton

The contentions of Professor Kinnicutt, set forth in the above paragraph, are supported by abundant evidence, and it constitutes as strong a statement in support of the vv^riter's position as he himself could ever hope to draw ; therefore the paragraph is submitted without discussion or amendment.

The analyses quoted in the following paragraphs are merely the chosen representatives of a great number that give the same testimony. They all show clearly the amount and condition of the nitrogenous matter, and can be used' to dififerentiate in some small degree between a comparatively stable and an unstable form of organic matter in water. But they show further that all those finely drawn distinctions by which we are supposed to determine whether or not such organic matter is of benign or dangerous origin are too precarious to be seriously considered. In every case it is easy to find a host of discrediting exceptions ; and when we go beyond the confines of New England and the country immediately there- about, and especially when we select our samples from the South or the Middle West or Far West, those exceptions become the rule.

That real man, the lamented friend of the most of those con- tributing to this volume, Dr. Thomas M. Drown, found not a few places in or near New England where his standards of interpreta- tion were useless. For example, many of us remember hearing him say that the polluted water of the Hudson above Poughkeepsie, N. Y., does not show upon sanitary analysis any traces of sewage matter. Yet neither he nor, it is believed, the most enthusiastic sup- porter of nitrogen determinations would accept that raw water as a beverage. In later years, not many months before Dr. Drown's death, the writer discussed with him the advisability of making an extended series of sanitary analyses upon the waters of the Lehigh River basin. Dr. Drown approved of a sanitary survey, but failed to see any promise in the analytical work. He ended his discussion of the matter by saying: "My long experience in this hne of work has impressed me with many doubts concerning its value."

The practice of making sanitary analyses and of judging the potability of a water from them has cost many lives. The cases are numerous and too well known to require discussion. In really competent hands such analyses do not usually produce serious

Futility of a Sanitary Water Analysis 49

results, because they are relegated into their proper place; but the supposedly competent hands are frequently brought to book. Let us review an instance.

In the memorable case of the State of Missouri vs. the State of Illinois and the Sanitary District of Chicago there was introduced into evidence the testimony of a professor of chemistry who qualified as an expert by relating all sorts of educational experience, both foreign and domestic. Cross-examination developed the following:

Q.: Taken in the abstract, without reference to anything else than the elements of pollution stated by you, to wit, free ammonia, 0.063, nitrites 0.002, albuminoid ammonia 0.552, nitrates 0.39, are those figures sufficient to warrant you in a con- clusion as to the potability of the water ?

A.: I think so.

Q. : What is your conclusion ?

A.: It is a potable water

Q.: Do you consider a water having the following constituents potable, namely, free ammonia 0.217, nitrites 0.013, albuminoid ammonia 0.676, nitrates 0.6?

A.: No, sir.

Q. : On what account ?

A.: Because the free ammonia has gone beyond 0.2, and the nitrites are up in the second place, whereas potable water should not have free ammonia very much above o. i; and, in fact, if the nitrites are measurable at all, we usually condemn the water;

Here is a man who gave two positive opinions concerning pota- bility of two waters, from the bare statement of the four nitrogen determinations. He did not think it necessary to take into account the other conventional statements. There are two lamentable features of this: first, he is teaching sanitary chemistry to students in a high- grade university; second, he is only one of a large number of persons similarly situated who are addicted to precisely the same absurdities.

It is anticipated that one of the principal objections made to the foregoing discussion will be that the examples given are exceptional cases, and that a far greater number of examples can be adduced which will support the standards of interpretation; that a few excep- tions do not, in science, destroy a theory, and that a great mass of data collected during past years should be accepted as the deter- minative basis. The cases presented for illustration are not excep- tional ones, nor, indeed, are they the best that might have been selec- ted for the purposes of this paper. If, however, we admit, for the

V

5© Marshall O. Leighton

purposes of argument, that they may be exceptional, the contentions of the writer are not damaged thereby. It should be remembered that in making sanitary analyses we are not developing a scientific theory that must stand or fall according to the weight of cumulative evidence for or against, but we are trying to determine whether or not the use of that water will cause sickness and death. This is a positive purpose ; and if it is admitted that there can be exceptions, even though they be few, the whole scheme of analytical procedure fails of that purpose. Exceptions are not' predestined, and in this case cannot be guided or defined. The chemist who calls a water ''good" upon the evidence presented by his nitrogen determinations has no means of knowing whether or not this water may be one of the exceptions. Supposing it be polluted like the Lake Champlain samples in Series "B," and a family or a community accepts the favorable opinion of the chemist and is stricken with typhoid fever — think you that that chemist will be justified by appearing before those bereaved relatives and reciting the fact that the great mass of evidence sustains certain bases of interpretation, and the scat- tering exceptions do not, from a scientific standpoint, destroy the integrity of the theory ? The most befitting remark at this junc- ture seems to be the old proverb: "A chain is no stronger than its weakest link."

After all, perhaps the strongest indictment of the sanitary analy- sis is that it is unnecessary for the purposes for which it is generally used. No one will question its value in sewage experiments, but when the purpose is to determine whether or not a water be potable, there are more satisfactory ways of solving the problem than by making the conventional grind of nitrogen determinations, even though it be admitted for the moment that those determinations fulfil all the great purposes claimed for them.

It may be accepted as axiomatic that no river, upon the drainage area of which there is any population, will furnish a water fit for domes- tic consumption in its raw state. That this shall hold true it is not necessary that the population shall be gathered into cities and be provided with sewerage systems. Rural population, even though widely scattered, is dangerous. Wherever people live along the banks of a stream there will always be dangerous pollution. Indeed, the

Futility of a Sanitary Water Analysis 51

natural drainage from occupied land is not always innocuous; but when this is combined with those direct and generally surreptitious pollutions, the effect is sometimes more acute than that produced by the everyday discharges from a city sewer. It will be necessary merely to recall the history of some of our classic typhoid epidemics to dem- onstrate this. The Plymouth, Pa,, epidemic was caused by a single focus of infection upon a sparsely settled drainage area. The New Haven epidemic arose from a similar cause, and upon a drainage area not only sparsely settled, but supposedly well protected. It is especially significant, too, that the Lowell and Lawrence epidemics did not have as their immediate cause the infected sewage from cities above on the Merrimack, but, as shown by Professor Sedgwick, from one or two incidental pollutions of Stony Brook. The same principles apply forcibly to the more recent epidemics at Butler, Pa., and Ithaca, N. Y. These remarkable instances illustrate the dangers of surface-water drainage from sparsely settled countries; it is obviously unnecessary to discuss similar dangers from water into which city sewage is poured. The question whether a river water will purify itself from such discharges in a given distance below a sewer outlet does not enter even remotely into this considera- tion, for, assuming that the sewage discharges would be purified, the incidental pollutions above a water intake would still constitute a grave danger. In the case of the Lowell and Lawrence epidemics, for example, perfect sewage purification at Manchester, Concord, and other cities on the Merrimack above the Lowell intake would not have prevented the scourge of typhoid. Therefore it is contended that, if we accept the principle that no surface-water draining from an inhabited area is safe in its raw state for domestic consumption, we shall err, if we err at all, upon the safe side, and there is no question that we shall save lives. What, then, is the necessity for analyzing river water for pollution, if we are all agreed that it must inevitably be polluted ?

Upland conserved supplies present a different phase of the ques- tion. If a drainage area contributing to a reservoir is in primeval condition with respect to population, it is generally admitted that the water must be wholesome. Why, then, make sanitar\- analyses to determine the presence of sewage ? If it be contended that this

f

52 Marshall O. Leighton

area may be subject to occasional malicious pollutions by visitors, etc., then the sanitary analysis does not offer any helpful solution. A single infectious intestinal discharge deposited directly in a reser- voir might readily cause a typhoid epidemic, but the organic matter would not, except under very fortunate circumstances, be detected by the nitrogen determinations; and it is well to reflect that, under the usual conditions, by the time the disease had made itself manifest and attention directed to that reservoir, the infection would have passed out of the reservoir.

If the drainage area above described does contain population, then the danger is always impending, and we may rest upon assump- tions, nearly, if not quite, as positive as those quoted above for river waters. It is quite significant in this connection to note that the Commission on Additional Water Supply of the City of New York made provision for filtration, although the upland areas proposed as new sources are sparsely settled. More recently we have read the opinions of our foremost authorities that the present Croton supply should be filtered. It is doubtful if any of those authorities would contend that the sole object of such filtration is to remove turbidity, color, and odor. It is therefore held that the sanitary analysis of upland conserved supplies is needless, because we can determine the danger by inspection far more readily and surely.

With reference to ground waters : We have interesting accounts of cases in which it is asserted that the condition of pollution was not detected by biological examination, but was revealed by sanitary analysis. If close consideration be given to the descriptions of premises that appear in these accounts, it will be seen that in every case (so far as the writer is informed) a careful man would have been justified in condemning those waters upon superficial examina- tion, and without regard to analysis. Take for illustration the case cited by Professor WiUiam P. Mason in a paper entitled, " Interpre- tation of a Water Examination," which appeared in Science^ Vol. 21, No. 539, pp. 648-53. It appears from this that there was a certain farmhouse in England, the residents of which had suffered severely from diphtheria and typhoid fever. Examination showed that the sewage discharged from the house entered into a dry-steyned cess- pool, without overflow, about four yards from the well, both sunk

Futility of a Sanitary Water Analysis 53

in gravel. In this case the chemical examination revealed the pres- ence of an excess of chlorides and nitrates while bacteriological investigation showed nothing which would cause suspicion. This instance is cited by Professor Mason as one in which "the danger signal was held out by the chemical side of the investigation alone." The writer is of the opinion that the "danger signal" was not the findings of the analysis, but the occurrence of disease in that residence. It should have caused an immediate examination of the premises, and such examination would have revealed the fact that the sewage was discharged into a dry-steyned cesspool, that the well was only jour yards away and that both were sunk in gravel. In the face of all our knowledge of the transmission of water-borne diseases, and in view of our decades of experience with infected wells, from historic Broad Street down to the present, why should any competent obser- ver, with the above related facts before him, find it necessary to fuss with an ammonia still or, for that matter, with a Petri dish? Taking a broad view of the subject of well supplies, we may safely exclude all wells in questionable places; and the careful observer can usually define such places.

All of the above discussion with reference to the needlessness of sanitary analyses, when other and more expeditious methods can be used, is based upon the temporar)^ admission that such anal- yses afford data whereby dangerous animal pollution can be dis- tinguished from harmless vegetable matter. If we return now to the original contention that standards of purity, bases of interpreta- tion, composition ratios, or by whatever name they may be called, are met with equal faith by the normal water and by the dilute sewage, and sum up the two lines of evidence, we have what the writer feels justified in regarding as an established case against the sanitary analysis as an index of dangerous water pollution.

V

THE VALUE OF PURE WATER*

George C. Whipple. PREFACE.

In order to estimate the relative value of waters which differ materially in quality, it is necessary to have some common denomi- nator. Nothing better for this purpose has been suggested than the dollar, which in this paper is made the basis of computation. By ascertaining what different characteristics of water cost the con- sumers, and by finding out how much consumers are willing to pay to avoid using waters which possess certain characteristics, an attempt has been made to secure a reasonable basis of comparison. The results of this initial study are here presented. They must not be taken too seriously at present, as some of the involved as- sumptions have not been established beyond doubt; and with the accumulation of certain data, necessary but not as yet obtainable, the results must be somewhat modified. Yet the general conclu- sions ought not to be far astray, and, from a study of the best data available, the writer believes that they err on the side of conserva- tism rather than on the opposite side. The suggested method of calculating the value of pure water seems to be one capable of being refined to a degree where its results will be of great practical value. The lines along which the accumulation of data is necessary in order to render the method reliable will be evident from a perusal of the text.

PURE AND WHOLESOME WATER.

To define the meaning of the expression "pure and wholesome water," which is so often found in water-supply contracts, would seem to be an easy matter, after all the study that has been given to the subject in recent years; but, although everyone knows in a general way what is implied by this expresssion, yet when it comes to framing a definition in positive scientific terms, the problem is not as easy as it seems. This is not because the chemist and the biolo- gist do not know what pure water is, but because water has so many

♦Received for publication February 17, 1906.

54

The Value of Pure Water 55

attributes which have to be taken into consideration, and because these attributes vary in importance in every instance. "Pure and wholesome water" is not a substance of absolute quality. Strictly speaking, pure water does not exist in nature; all natural waters contain substances either in solution or suspension; and in propor- tion as these substances are present, and in proportion as they are objectionable in character, the water is impure. Definitions of pure and wholesome water, therefore, generally state what foreign sub- stances shall not be present, or in what amounts they are permis- sible, instead of defining the positive qualities which the water shall possess.

Unquestionably the term "pure and wholesome water," as ordi- narily used, relates to water intended to be used for drinking. Such a water must be free from all poisonous substances, as the salts of lead ; it must be free from bacteria or other organisms liable to cause disease, such as the bacilli of typhoid fever or dysentery; it must also be free from bacteria of fecal origin, such at B. coli. In other words, the water must be free from poisonous substances, from infection, and even from contamination,* Besides this, it must be practically clear, colorless, odorless and reasonably free from objection- able chemical salts in solution and from microscopic organisms in suspension. Moreover, it must be well aerated. Color, turbidity, odor, dissolved salts, etc., may be permissible to a small degree with- out throwing the water outside of the definition of pure and whole- some waters. In these minor matters local standards govern up to a certain point, and it is in regard to them that differences in the judgment and experience of analysts lead to different classifications.

When it comes to using water for other purposes than for drinking, other attributes have to be considered. Hardness makes a water troublesome to wash with and to use in boilers; iron makes trouble in the laundry; chlorine corrodes pipes and makes work for the plumb- ers; the presence of the carbonates and sulphates of lime and mag- nesia affects the paper-maker, the brewer, the tanner, the dyer, the bleacher; soda causes a locomotive boiler to foam, and affects the use of the water for irrigation. All of these constituents, and others which

♦By this term is meant pollution with fecal matter. Contamination must be considered as potential infection.

56 George C. Whipple

are not named, have to be taken into consideration in connection with a public water supply, w^hich may be put to any of these uses.

If it is a difficult matter to define a pure and wholesome water in strict scientific terms, it is still more difficult to compare waters which differ in purity on any reasonable basis; and yet this often has to be done. Given two water sources equally available to a city for pur- poses of supply, both safe to drink, but one high-colored and soft, the other colorless and hard — which is the better selection ? A water- works plant is to be appraised : structurally the system is a good one but the quality of the water is unsatisfactory because of its excessive color or turbidity — how much should be deducted from the value of the works because of the bad quality of the water ? The water- works owned by a private company are to be purchased by the city ; the city has a high typhoid fever death-rate due unquestionably to the water supply — how much less should the city pay because of that fact ? A city in the West is using turbid river water — how much can it afford to pay to filter it ? A city in New England is using a water so heavily laden with Anabaena that it is nauseous to drink — how much can the city afford to pay to procure a new supply ? These are all practical, everyday questions which deserve answers based on scien- tific data.

In valuation cases, where the quality of the water supply has been unsatisfactory, the cost of filtration, or other appropriate method of purification, has been sometimes taken as a measure of the inferior quality of the water, and this amount deducted from the value of the works. In case filtration was impractical, or more expensive than securing a supply from a new source, the additional cost of such new supply has been sometimes taken as a measure of the inferior quality of the works and the amount deducted from the value of the works. Both of these methods are similar in that they contemplate the sub- stitution of a satisfactory water for one not satisfactory.

Another method of measuring the depreciation applicable to a water-works plant because of an inferior quality of the supply would be to ascertain what the use of the impure water has cost the consumers, compared with what a pure and satisfactory water would have cost them. This method has not been used in practice, but it seems to be a reasonable one, and one which would be of more general

The Value of Pure Water 57

application than the preceding, if the data upon which it is based could be accurately determined. Unfortunately, this is not the case in many instances, but by the use of certain generalized data and assumptions, results may be secured which are of considerable use in comparing the value of waters different in quality.

The qualities of a public water supply which most affect the ordi- nary consumer are :

1. Its sanitary quality; that is, its liability of infection with disease germs or substances deleterious to health.

2. Its general attractiveness, or lack of attractiveness, as a drinking- water.

3. Its hardness, so far as this relates to the use of soap in the household.

4. Its temperature, so far as this relates to drinking. Characteristics which affect industrial uses are too much a matter

of local concern to be taken into account in a general discussion, although they are by no means of small account, and in some com- munities their importance might control. The qualities selected are to be considered as illustrative of the method rather than as a com- plete exposition of it.

The problem is to express these four characteristics in terms of dollars and cents to the consumer. The financial standard is cer- tainly not the highest one for judging the quality of a water supply when the public health is concerned ; human life cannot be estimated in gold dollars, and the smell of unsavory water to a thirsty man cannot be reckoned in dimes; nevertheless, the financial basis is a convenient one, and one necessarily involved in all questions which relate to public utilities.

SANITARY QUALITIES.

If the water under consideration has been used for a considerable time, the typhoid fever death-rate of the community will fairly well represent the sanitary quality of the water supply. It will not tell the whole story, but in most cases it will not lead far astray. In order to reduce this to a financial basis, it is necessary to make several assumptions.

The financial value of a human life is generally taken as $5,000,

58

George C. Whipple

but according to Leighton' it varies at different ages from $i,ooo to $7,000, as shown by Table i. It so happens that persons are most susceptible to typhoid fever near the age when their life-value is considered greatest. By combining the life-value at different ages with the age distribution of persons dying of typhoid fever, the resulting average value of persons dying from typhoid fever is found to be $4,634, which is very close to the figure ordinarily used.

The percentage mortality of typhoid fever patients is sometimes stated as 10 per cent; that is, ten cases for every death. Figures of this character are most often based on hospital records, and mild cases do not generally reach the hospitals. Studies of recent typhoid epi- demics indicate that 15 to 18 cases for each death would be nearer the truth. The expense of medical treatment, nursing, and medicine, the loss of wages for a month or more, together with other attending expenses and inconveniences, would doubtless aggregate at least $100 per case, or $1,000 for the 10 cases corresponding to one death. If the estimate of $100 is considered too large, it may be answered that the excess is more than offset by the fact that there are more often from 15 to 18 cases for each death than there are 10. It may be fairly assumed, therefore, that $6,000 is a very moderate estimate of the financial loss to the community from typhoid fever for each death from that disease.

TABLE I.

Age o- s years

s-° ::

10-15

15-20 "

20-2S "

25-30 "

30-35 '

35-4° ;;

40-45

4S-50 ::

50-55

SS-60 "

60-65 "

65-70 "

70-

Total

Estimated Value of Human Life

$1. 2 2 3 5' 7 7 6

5 5 4 4 2 I I

500 300 500 000 000 500 ,000 000 500 ,000 Soo 500 ,000 ,000 ,000

Per Cent of Deaths from Typhoid Fever

5

7

13

16

13 9 8

S 4 3 2 2 I I

Product of Columns 2 and 3

$ 7-5 10

13.570

18,000

39.300

83.500

99,100

60,300

48,000

30,900

20,000

15,000

11,700

4,200

1.500

1,900

$463,480

Average value of life of persons dying from typhoid fever, $4,634. 'M. O. Leighton, Popular Science Monthly, January, 1902.

The Value of Pure Water

59

TABLE 2.

Effect of FatRAXiON on Death-Rates at Albany, N. Y., and a Comparison with Troy, N. Y.,

Where the Water Was Not Filtered.

Death-Rates per 100,000

1804-98, before Filtration at Albany

I 900- I 004

after Filtration

at Albany

Difference

Per Cent Re- duction of Death Rates

Albany
Typhoid fever	104 125 606 2.264	26 53 309 1,868	78 72 297 378	75
Diarrheal diseases	57
Children under 5 years	49
Total deaths	17

Troy
Typhoid fever	57 116 531 2,157	57 102 2,028	0 14 06 120	0
Diarrheal diseases	12
Children under 5 years	18
Total deaths ...	6

Remark: Filtered water was introduced into Albany in 1899. of Troy has remained practically unchanged.

The water supply

Typhoid fever is by no means the only disease transmitted by contaminated water. Dysentery and various other diarrheal diseases precede it or follow in its train, and in most instances these are prob- ably due to the same general sources of contamination as those w^hich caused the typhoid fever, although, of course, to different specific infections. The reduction of the typhoid fever death-rate following the substitution of a pure water for a contaminated water is often accompanied by a drop in the death-rate from other diseases. Thus, if the five years before and after filtered water was introduced into Albany, N. Y., are compared, it will be seen that the reductions in deaths from general diarrheal diseases and the deaths of children under five years of age were much greater than in the case of typhoid fever. There was also a reduction in malaria, but this probably represents faulty diagnosis of typhoid fever cases before the introduc- tion of the filters, rather than a real reduction of malaria. That the reduction of infant mortality and deaths from diarrheal diseases was not due to other conditions seems probable from the fact that in the neighboring city of Troy, where the water supply was not changed, there was no such diminution during the same period. (Sec Table 2.)

Hazen, in his paper on "Purification of Water in America," read

6o George C. Whipple

at the International Engineering Congress at St. Louis, called atten- tion to this same fact, that after the change from an impure to a pure supply of water the general death-rate of certain communities investi- gated fell by an amount considerably greater than that resulting from typhoid fever alone — indicating either that certain other infectious diseases were reduced more than typhoid fever, or that the general health tone of the community had been improved. Thus, for five cities where the water supply had been radically improved he found :

Per icx3,ooo Reduction in total death-rate in five cities with the introduction of a pure water

supply 440

Normal reduction due to general improved sanitary conditions, computed from

average of cities similarly situated, but with no radical change in water supply 137

Difference, being decrease in death-rate attributable to change in water supply . 303 Of this, the reduction in deaths from typhoid fever was 71

Leaving deaths from other causes attributable to change in water supply . . . 232

From these facts it is evident that to place the financial loss to a community as $6,000 for each death from typhoid fever due to the public water supply is to use too low a figure. It probably ought to be several times as high ; but recognizing the lower financial value placed on the lives of infants, and the less serious character of the other dis- eases, and wishing to be as conservative as possible, for the reason that typhoid fever is not entirely a water-borne disease, $10,000 per typhoid death has been used in the calculation which follows.

Since typhoid fever is a disease which may be transmitted in other ways than by the water (as, for instance, by milk, shell-fish, or flies), it is necessary to allow a certain death-rate for these other causes, for even in a city where the water supply is perfect there may still be some typhoid fever. To establish this "normal"* is a difficult matter, but for purposes of calculation we may assume it to be determined and represent it by the letter N.

If we assume that all typhoid fever in excess of N is due to the water supply, and if we assume that the daily consumption of water is 100 gallons per capita, then letting T equal the typhoid fever death- rate per 100,000 —

(T—N) 10,000 = loss to the community in dollars for 365 X 100 X

11 r ^ T> {T-N)i,ooo 100,000 gallons of water, or D= 7 ■ =2.75(7 —A'),

♦This term "normal" must not be assumed to mean necessary typhoid.

The Value of Pure Water

6i

where D stands for the loss in dollars per million gallons of water used.

Suppose the average typhoid fever death-rate for a community which has a somewhat polluted water supply has averaged 43 per 100,000 for a period of five years, and suppose that for this place the value of A'' is estimated as 15, then —

Z) = 2.75 (43—15) ~ $76.72 if the per capita consumption is 100 gallons. If the consumption per capita is 115 gallons, D would be W^ of $76.72, or $66.71; if it were 63 gallons per capita, then D would equal ^^^ of $76. 72, or $121 . 77.

The value of N must be naturally subject to local variation, and in order to obtain an idea as to its probable value, a compilation of typhoid fever death-rates has been made for cities and towns in differ- ent parts of the country which use ground waters or filtered waters — that is, waters which may be considered as free from contamination.

The following is a generalized summary of them :

TABLE 3. Typhoid Fever Death-Rates m Cities and Towns Which Have Ground- Water Supplies.

State

Maine

Massachusetts Connecticut . . New York . . . New Jersey . . Pennsylvania. Ohio

Number of Cities

and Towns

Averaged

2 23

4 I3 10

s

22

Number of Years Averaged

Average Typhoid

Fever Death- Rate

per 100.000

6.4 15-8

9-5 24.7 20.5 31.8 32.4

There is reason to believe that the higher rates given above do not correctly represent the situation, because in some instances the ground water was supplemented by the occasional use of water which may have been polluted. Proximity to a large city where the water supply is contaminated was also responsible for some of the high figures; so also was the absence of sewerage systems. Nevertheless, there seems to be a slight tendency for the typhoid fever rates to increase in the United States from north toward the south in those places where the water supply is reasonably safe. There are some exceptions to the increase southward, however. Thus, in Camden, N. J., which is supplied with a pure ground water, the typhoid rate in 1901 was only 12, and 20 in 1902.

62 George C. Whipple

In Fuertes' book on Water and the Public Health sl diagram is given showing that the typhoid fever death-rates in cities supphed with ground water vary from 5 to 32 per 100,000 in America, and from 6 to T)T, per 100,000 in Europe, the average being about 18 in America and 19 in Europe. It is shown also that the death-rates from cities supphed with fiUered water vary from 4 to 20 in America, and from 4 to 20 in Europe, the average being 12 in both cases. Recent Ameri- can data for cities supphed with fiUered water show that the rates are somewhat higher than these, the average being somewhat less than 20.

Taking into consideration the best available data, it seems reason- able to place the general value of A^ somewhere between 10 and 25 per 100,000, with the most probable average value as 20, which figure may be used in the equation where local sanitary conditions are unknown. The value of N, however, should be varied where there is reason for doing so. Where the sanitary conditions are good 15 may be taken as a fair value. In New England it might be placed lower than in regions south of the glacial drift ; in cities near the seaboard, where there is a large consumption of oysters consumed fresh from the layings, the value of N might be higher than in inland cities, where the oyster consumption is small and where fattened oysters are not used as freely; in cities where there are cess-pools, but no sewers, the value of N would naturally be higher than in cities well provided with sewers.

It may be reasonably expected that, as time goes on, the value of N will gradually fall, because of a general decrease of typhoid fever in the country at large, and a consequent diminution of the number of foci of infection. Statistics for twelve states, including all the New England states. New York, New Jersey, Maryland, California, Min- nesota, and Michigan, show that during the last quarter of a century the general typhoid fever death-rate has fallen as follows:

TABLE 4

Average Typhoid Fever Death- Rate per Year 100,000

1880 55

1885 46

1890 36

1895 28

1900 23

1905 21

The Value of Pure Water

63

ATTRACTIVENESS.

The analytical determinations which relate to the general attrac- tiveness of a water are those of taste, odor, color, turbidity, and sedi- ment. As these quantities increase in amount, the water becomes less attractive for drinking purposes, until finally a point is reached where people refuse to drink it. In order to use these results in a practical way, it is necessary to combine them so as to obtain a single value for the physical characteristics or, as they say abroad, for the "organoleptic" quality of the water. An attempt has been made by the author to obtain what may be termed an esthetic rating of the water, and the result is shown in the accompanying diagram.

I X <i ij

Tu^eiD/Tf jif^c coj-off (^jatrrj ^jr^ AJ/j-^'On/J

This diagram, it should be said, is based almost entirely upon esti- mates and very little upon statistical data. It rests upon the assump- tion that people differ in their sensibilities, or their esthetic feelings as to the use of water. Some persons are much more fastidious than others in regard to what they drink. A water which would be shunned by one person, even though he were thirsty, might be taken by another with apparent relish. As a rule, people are more fastidious about the odor of water and the amount of coarse sediment which it contains than they are about its color and turbidity. This is perhaps natural, as a bad odor suggests decay, and decay is instinctively repugnant.

64 George C. Whipple

Often, however, people do not discriminate between odors which are due to decomposition and those which are not. Habit and associa- tion have much to do with a person's views as to the attractiveness of water. In New England, where the clear trout brooks run with what Thoreau called "meadow tea," few people object to a moderate amount of color, while they do object to a water which is very turbid. In the Middle West, where all the streams are muddy, it is the un- known colored waters which are disliked. People who are accus- tomed to well water object to both color and turbidity. With most people a fine turbidity, such as is produced by minute clay particles, is less a subject of complaint than an equal turbidity produced by comparatively coarse sediment. In the diagram an attempt has been made to reconcile these different points of view so as to put them, as weU as may be, on the same footing. In this connection several series of comparisons were made.* Turbid waters were viewed through the eyes of a group of western people, who made some comparisons with color and turbid waters, while colored waters were viewed through the eyes of a group of students in New York, and vice versa.

The abscissae of the diagram represent turbidity, color, and odor, as given in the ordinary water analysis. f The ordinates represent the "per cent of objecting consumers." By this is meant the proportion of the water-takers who would ordinarily choose not to drink the water because of the quality indicated by the curve, or who would buy spring water, or bottled water, rather than use the public supply, if they could afford to do so. This number would increase, of course, as the general attractiveness of the water decreased. From the curves one may calculate what may be called the esthetic deficiency of the water by adding together the per cents of objecting consumers for color, turbidity, and odor. If the esthetic deficiency equals loo, it indicates that the water is of such a character that everyone would object to it, and figures in excess of loo only emphasize its objectionable character.

It will be seen from the diagram that when the color of water is less than 20, or the turbidity less than 5, only one person in ten would object to it, but when the turbidity or color is 100, one-half of the

♦Acknowledgments are due to Mr. J. W. Ellms, of Cindnnati, Ohio, and Mr. Andrew Mayer, Jr., of Brooklyn, N. Y.

tSee " Report of Committee on Standard Methods of Water Analysis, American Public Health Asso- ciation," Supplement No. i. Journal of Infectious Diseases, May, 1905.

The Value of Pure Water 65

people would object to it. It may be thought that this proportion is too low, but it must be remembered that colored waters are invariably accompanied by a vegetable odor and often by a slight turbidity, and that it is the sum of the several quantities which determines the esthe- tic rating.

Experience has shown that objection to color varies directly with its amount ; consequently this curve has been plotted from the equa- tion, pc = — , i. e., a straight line, where pc stands for the per cent of

objecting consumers, and c for the color.

In the case of turbidity, however, small amounts count for more, relatively, than larger amounts. The equation for the turbidity curve has been taken, therefore, a,s pi = $V t, where t stands for the turbidity.

With odor, however, the opposite condition prevails; faint odors count for little, but distinct and decided odors cause much more com- plaint. Consequently, the per cent of objecting consumers has been made to vary as the square of the intensity of the odor expressed according to the standard numerical scale. The quality of the odor makes quite as much difference as its intensity, and for that reason three curves have been plotted, one representing vegetable or pondy odors (Oj,), one representing odors due to decomposition (Oj), and one representing the aromatic grassy and fishy odors due to micro- scopic organisms (Og). These curves are plotted from the following equations :

p.=5o:,

Po=20l,

in which O^, Od, and Oy stand for the intensity of the three groups of odors mentioned.

These curves represent somewhat imperfectly our present ideas as to the relative effects of color, turbidity, and odor; and on further study they are likely to be considerably modified.

It is a well-know^n fact that in cities which are supplied with water which is not attractive for drinking purposes, large quantities of spring water and distilled water are sold, and that consumers go to much expense in the purchase of house filters in order to improve the quality of the water furnished by the city mains. It is fair to assume

66 George C. Whipple

that in any community the amount of money expended for bottled water and house filters will vary in a general way according to the attractiveness of the water, although there is no doubt that the presence of typhoid fever in the community, or the fear that the water is con- taminated, will greatly increase the use of auxiliary supplies for drink- ing. For purposes of calculation it may be assumed that the diagram just described represents this tendency to use vended waters, and that each "objecting consumer" would go to the expense of buying spring water or putting in a house filter, if he could afford it. It may be argued, also, that the poor consumer who may be unable to do this is as much entitled to satisfactory water as is the well-to-do consumer.

From a study of price-lists of spring waters sold in New York and other cities it has been found that the ordinary wholesale price of spring water is seldom more than lo cents a gallon. In some places it is as low as i cent. The average is about 5 cents. To filter water through house filters costs less, but generally it is less satisfactory.

As a convenient figure for calculation, and as a most conservative one for general use, a cost of i cent per gallon to the ordinary con- sumer for an auxiliary supply of drinking-water (either spring water or well-filtered water) has been taken. In cities where the cost of procuring and distributing bottled water exceeds i cent per gallon, as it does in such a city as New York for example, this should be taken into account in making local use of the data. For the illustrative pur- poses of the present paper, and for general comparisons, the figure mentioned will serve as a satisfactory basis. The average person drinks about i . 5 quarts of water per day, and therefore one-fifth cent per capita daily may be taken as a reasonable figure for the cost of an auxiliary supply. If the entire population used such a supply, and if the daily consumption of the public water supply were 100 gallons per capita, then one-fifth cent per hundred gallons, or $20 per million gallons, would represent the loss to the consumers due to an imperfect water supply which had an esthetic deficiency of 100. If the esthetic deficiency were less than 100, say 37, then the loss to the consumer would be yVt of $20, or $7 . 40 per million gallons. In other words, the figure for the esthetic deficiency divided by 5 gives the financial depreciation value of the water supply in dollars per million gallons,

or// = 20 .

100

The Value of Pure Water 67

Example: Suppose the turbidity of a water is 3, its color 65, and its odor 2f (that is, faintly fishy), because of the presence of micro-

I 2 -|- '22 "i" 20

scopic organisms; then D = 20 =$12.80; that is, the depre- ciation value of the water, because of its unsatisfactory physical qualities, amounts to $12 .80 per million gallons.

HARDNESS.

The point at which a water becomes objectionably hard has never been exactly defined. Standards of hardness vary in different parts of the country. The ordinary person washing his hands considers the water soft if the soap will quickly produce a suds without curdling. A hardness of 10 parts per million is practically unnotice- able, and it requires a hardness of 20 or 30 parts per million to produce ''curdling." Waters which have a hardness below 25 parts per million seldom cause much complaint, but when the hardness rises above 50 the water is well entitled to the appellation "hard," and above 100 it may be called very hard. In some parts of the country hardnesses of 200 or 300 are observed; these may be termed "excessive."

In 1903 a number of experiments were made by the writer to deter- mine the effect of various degrees of hardness on the amount of soap used in washing the hands, in bathing, and in general household uses. As a result of these experiments it was found that one pound of the average soap as used in the household would soften 167 gallons of water which had a hardness of 20 parts per million. This was equiva- lent to about three tons of soap per million gallons, which at a cost of 5 cents per pound, would amount to $300 per million gallons. It was found also that for every increase of i part per million of hardness the cost of soap increased about $10 per million gallons of water softened.

All of the water used by a community is not completely softened. The number of gallons per capita per day completely softened has been estimated by different authorities all the way from i to 10. It will certainly be a conservative estimate to assume that one gallon per capita is thus softened. On this basis the depreciation value of

water, on account of its hardness, is D = — , in which H equals

' ' 10 ' *

the hardness of the water in parts per million, and D the depreciativc value in dollars per million gallons.

68 George C. Whipple

Example : Assume the total hardness of a water to be 79 parts per

79 million ; then Z) = — = $7 . 90 per million gallons.

This takes into account only the cost of soap used for domestic purposes, and does not include the incidental losses and inconveniences attendant upon the use of hard water in the household. These, if they could be expressed in terms of dollars and cents, would probably more than equal the cost of soap; therefore the above figures err on the side of conservatism.

TEMPERATURE.

Everyone knows that warm water is unpalatable. When the tem- perature rises above 60° F., people do not like to drink it without cooling. The relation between the temperature of the water and the per cent of objecting consumers may be represented by a curve based

on the equation /> = , in which p equals the per cent of

objecting consumers, and d equals the temperature of the water in Fahrenheit degrees. According to this curve, no one would object to drink a water which had a temperature of 45°, half the people would object at 66°, and all would object at 75°. If it is assumed that it takes one-half pound of ice per capita daily to cool the water used for drinking during four months in the year, and that ice costs 30 cents per 100 pounds, then the depreciation value due to temperature would be equivalent to $5 per million gallons of public supply for 100 per cent of objecting consumers, assuming the

per capita consumption to be 100 gallons daily, otD = — X $5 = ^^ — ^— ^

in dollars per million gallons, in which d = the average temperature during the four warmest months of the year. This may be considered as the depreciation value due to temperature. The temperature of ground waters seldom rises above 60° in the house taps even in summer, and in cities supplied with ground water a large propor- tion of the consumers do not use ice. Surface waters, on the other hand, in the latitude of New York, generally maintain a temperature of 60° or more at the house taps for at least four months of the year. The temperature factor is an important one in many cases,

The Value of Pure Water 69

but it need not be used except when comparing surface waters with ground waters.

In a similar way it might be possible to calculate the reduced value of a water due to other objectionable characteristics, such as the presence of large amounts of iron or chlorine. Except in special cases, these would not be as important as the more obvious qualities above described, and they need not be considered in this discussion.

SUMMARY OF PRINCIPAL FORMULAE.

Depreciation due to sanitary quality —

I. D=^2.ys(T-N).

Depreciation due to physical characteristics —

Pc + P^+Po

2. D = 20-

100

c

p, = Si^t

po=20i+s.so:i+soi.

Depreciation due to hardness —

^ 10

Depreciation due to temperature —

4. D = — , m which —

180 '

D = the depreciation value in dollars per million gallons;

r = typhoid fever death-rate per 100,000;

iV = typhoid fever death-rate assumed to be due to causes

other than water, and which may be ordinarily taken as

20 per 100,000; pc = peT cent of consumers who object to the color of the

water; />/ = per cent of consumers who object to the turbidity of the water;

Po = 'peT cent of consumers who object to the odor of the water; c = color reading; / = turbidity reading;

70 George C. Whipple

0^ = odors due to vegetable matter, expressed according to standard numerical scale;

0(i = odors due to decomposition, expressed according to standard numerical scale;

Oo = odors due to microscopic organisms, expressed accord- ing to standard numerical scale;

H = hardness of Water in parts per million ;

d = average temperature of water during four warmest months.

APPLICATION OF THE FORMULA.

It now remains to apply the principles above set forth to actual cases and see to what conclusions they lead.

effect of contamination.

The average death-rate from typhoid fever in American cities which have more than 30,000 inhabitants is about 35 per 100,000. Applying formula (i), and assuming a value of 20 for A^, then

^=2.75(35-2o)=$4i.25; that is, the average depreciation value of the water supplies of our American cities, taken as a whole, is $41 . 25 per million gallons because of their unsanitary quality, or about $15,000 per annum for each mil- lion gallons a day of supply.

The above figure takes into account both good and bad supplies. The average typhoid fever death-rate in those cities which have rea- sonably good water supplies may be taken in round numbers as about 20, while in those cities which have supplies more or less contaminated it varies from this up to 40 or 60. In some of the worst cases it is more than 100 per 100,000. In Pittsburg, for example, the typhoid death-rate for several years has averaged 120. Here, according to formula (i), D = 2.75 (120— 20) =$275 per million gallons. This is figured, however, on a per capita water consumption of 100 gallons a day. The actual consumption is about 250 gallons per capita per day; hence D should be taken as ^^^ of $275, or $110 per million gallons. Each million gallons of polluted Allegheny River water pumped to Pittsburg has therefore reduced the vital assets of the com- munity by $110. This, for a population of 350,000, amounts to $3,850,000 per year — a sum enormously greater than the cost of making the water pure.

The Value of Pure Water

71

Classifying water supplies according to their source, the following will give a general idea as to the depreciation value of various types of water from the sanitary standpoint, based on general average typhoid fever death-rates :

Charactes of Water Supply.

Depreciation Value in

Dollars per Million

Gallons

I.

So. 00

$0.00

$ 0.00 to $ 15.00

Ground waters, except in cases where pollution is excessive, or where wells are driven in rock or soil abounding in fissures .

2. Filtered waters (assuming modern methods of construction and operation),

3. Surface waters —

a) Ordinar>' upland waters, with insignificant contamination .

b) Shghtly contaminated waters, with good storage in lakes or large reservoirs 10.00 to 50.00

c) River waters, slightly contaminated, little or no storage . 25.00 to 100.00

d) River waters, much contaminated, little or no storage . . 50.00 to 200.00

EFFECT OF TURBIDITY, COLOR, AND ODOR.

It has been shown that the esthetic deficiency of water depends upon three variable characteristics, which may have many different combinations; consequently, it is difficult to classify the water sup-

TABLE 5. Examples of Waters wrrH Different Physical Characteri.stics.

City

Source of Supply

Turbid- ity

Color

Odor

Per Cent of Ob- jecting Con-

Portland, Me

Boston, Mass

Cleveland, Ohio. . Worcester, Mass . . New York City. . . Brooklyn, N.Y...

Jersey City, N. J.. Waterlown, N. V. Springfield, Mass.

Bangor, Me

Pittsburgh, Pa. . . Philadelphia, Pa.. St. Louis, Mo. . . .

Lake Sebago

Sudbury and Nashua Rivers

Lake Erie

Storage Reservoirs

Croton River

Ponds and driven wells on Long

Island

Rockaway River

Black River

Ludlow Reservoir

Penobscot River

Allegheny River

Schuylkill River

Mississippi River

Depreci-

ation\'al-

ue per

Million

Gals.

ground waters.
Camden, N. J.. . .	Driven wells	0 0 0	1 0 10	0 0 0	0 0 5	0.00
Flatbush, L. I. . . .	Driven wells	o.oo
Lowell, Mass	Driven wells	1 .00

surface waters.

I	IS	2V	20
3	2S	2V	30
18	5	I.SV	30
2	30	y"	40
4	20	3"	SS
3	13	15^	^t
4	32	2V \g	38
6	70	3"	SS
S	27	AS	104
6	6S	T,v im	so
64	30	iv zm	87
ISO	10	3V 2m	102
200	30	3V2m	127

$ 4.00 6.00 6.00

8.00 II .00

7.20

7 60 II .00 20.80 11.80 17.40 20.40 25.40

Some of the above figures do not represent present conditions. For example, Watertown, N. Y., now has filtered water; St. Louis uses a chemically treated water; etc.

72

George C. Whipple

plies of the country on this basis. For this reason the few typical examples given in Table 5 may be more instructive than any attempt at a general classification.

It will be seen from the above figures that, while the general attractiveness of a water is of less importance than its sanitary quality, yet it is by no means insignificant. For instance, such a water as that now supplied to New York City from the Croton River has a depreciation value of $11 per million gallons, or nearly a million and a half dollars a year for a daily supply of 350 million gallons. At 4 per cent this represents the interest on about $35,000,000, a sum several times as large as the cost of filtration. An algae-laden water like that of Ludlow Reservoir at Springfield, Mass., has a depreciation value of more than $20 per million gallons, because of its odor and turbidity. A colored water like that of the Black River at Watertown before filtration has a depreciation value of $11, while a turbid water like that of the Mississippi River at St. Louis gives $25.

In most surface waters the physical characteristics vary greatly at different times of the year. During the spring and fall, for instance, the color and turbidities may be high on account of rains, while during the summer the water may have bad odors due to microscopic organ- isms. The depreciation value of a certain reservoir water, calculated as above described, serves well to show this seasonal variation, as illustrated by the following figures :

TABLE 6.

Seasonal Variation in the Depreciation Value of a Surface Water Due to Seasonal Changes in Turbidity, Color, and Odor.

Month

January

February

March

AprU

May

June

July

August

September

October

November

December

Average

Turbid-	Color
ity
6	25
8	28
7	27
5	22
8	25
7	30
4	22
4	25
3	30
4	28
3	26
4	25

Odor

(3^ o (3V o (,3V o (3V 2

(3V I (3V 1 (3^ o (3^ o

3v 3v 3v 3v +

5 Org. 0.3m 3m

5 Org. osm Org. 0.5OT

Org. 0.5m Org. o.sm

3m

3"«

Per Cent Objecting Consumers

44 47 45 40

49 48 43 62

63 49 40 42

Depreciation of

Value per Million Gals.

$ 8.80 9.40

9. GO 8.00

9.80

9.60

8.60

12.40

12 .60 9.80 8.00 8.40

$ 9-53

The Value of Pure Water

73

EFFECT OF HARDNESS.

The waters of New England are comparatively soft, although in some instances the ground waters are hard. In the Middle West, on the contrary, most of the surface waters are quite hard, and in some cases the hardness is excessive. The following figures serve to give an idea of the range in the depreciation value of waters due to hard- ness.

TABLE 7.

State	City or Town	Source of Supply	Total Hardness (Parts per Mill.)	Deprecia- tion Value per Million Gals.
Maine	Augusta WaterviUe	Kennebec River Messalonskee River	20 15 12 33 SO 40 64 191 179 200 335 578 215 243	S 2.00 2 ^0
Massachusetts	Boston	Sudbury and Na,shua Rivers. . Storage Reservoir
	Cambridge	3 30 5 00 A 00
II	Pittsfield	Storage Reservoir
New York	New York	Croton River
	Albany	Hudson River.	6.40 19 10
«i	Oswego	Oswego River
Pennsylvania	Philadelphia	Schuylkill River	17 90
Ohio	Toledo	Maumee River
	Columbus	Scioto River	33-50 33-50 21.52 24 30 1
"	Warren	Mahoning River . . . .
England		Chelsea Company East London Company
	London

EFFECT OF FILTRATION.

Sanitary quality. — The following figures show to what extent the sanitary value of a polluted public water supply is increased by an efficient system of filtration :

Laurence, Mass. —

Water supply, Merrimack River, filtered by a slow sand filter.

Population 70,000.

Water consumption, 40 gallons per capita daily.

Before filtration the typhoid fever death-rate was 121 per 100,000; since then it has been 26.

Before filtration 2^ = 2.75 (121 — 20) X W =$693.

After filtration Z? = 2.75 (26-20) X W =$41-

Increase in sanitary value = $693 — $41 =$652 per million gallons, or $665,000 per year, or $9.50 per year per capita. Albany, N. Y. —

Water supply, Hudson River, filtered by sand filter.

Population, 95,000.

Water consumption, 165 gallons per capita daily.

Before filtration the typhoid fever death-rate was 104 per 100,000; since then it has been 26.

Before filtration Z) = 2. 75(104— 2o)X}§8 =$140.

After filtration D = 2.-j$ (26-20) Xi?g = $10.

74 George C. Whipple

Increase in sanitary value = $140— $io = $130 per million gallons, or $450,000

per year, or $4 . 75 per capita per year. Binghamton, N. Y. —

Water supply, Susquehanna River, filtered by a mechanical filter.

Population, 42,000 (approximately).

Water consumption, 160 gallons per capita daily.

Typhoid fever death-rate before filtration, 49; after filtration, 11 per 100,000.

Before filtration D = 2.-js (49- n) Xifg =$65.

After filtration Z) = 2. 75(11 — 11) Xj-n = o.

Increase in sanitary value = $65.00 per million gallons, or $160,000 per year,

or $3.80 per capita per year. Watertown, N. Y. —

Water supply. Black River filtered by mechanical filter.

Population, 25,500 (approximately).

Water consumption, 160 gallons per capita daily.

Typhoid fever death-rate before filtration, 68 per 100,000; after filtration, 19.5.

Before filtration D = 2.75 (68-20) XTgu = $82. 50.

After filtration ^ = 2.75 (20— 20) X VV =o-

Increase in sanitary value = $82. 50 per million gallons, or $120,000 per year,

or $4.75 per capita per year.

Illustrations like the above might be multiplied, but the four cases selected are sufficient to illustrate the general fact. It is easily seen from them that the filtration of a polluted public water supply increases to a very great extent the vital assets of a community, and the increase in most cases is many times greater than the cost of constructing and operating the works. Money paid to the doctor, the apothecary, and the undertaker is not, in one sense, a loss to a community, as it is merely a transference of money from one man's pocket to another's, but in the broader sense any loss of productive capacity or any unnecessary expenditure is a loss. Deaths from typhoid fever and from other diseases, however, represent a very material loss of the productive capacity of a community, and consequently a decrease in what may be termed the "vital assets." In the case of the city of Albany, for instance, the increased worth of the water to the city, because of its efficient filtration, amounts to $475,000 per year, of which at least $350,000 may be considered as a real increase in the vital assets of the city.

If in the formula D=$2.js (T-N) we let T-N = i, then D = $2.75; that is, a decrease in the typhoid fever death-rate of i per 100,000 causes an increase in the vital assets of the city of $2.75 for each million gallons of the public water supply (assuming this to be

The Value of Pure Water

75

loo gallons per capita), or $o.io per capita per year for each unit reduction of the typhoid fever death-rate per 100,000. In other words the decrease in the typhoid death-rate per 100,000 divided by 10 gives the increased vital assets of the community in dollars per capita per year. Thus in the case of Albany, above given, the reduction in the typhoid fever death-rate w^as 78 per 100,000. On the basis of 10 cents per capita per unit decrease, this would amount to $0. 10 X 78X95,000 = $741,000 per year, assuming a per capita consumption of 100 gallons daily, or $450,000 for a per capita consumption of 165 gallons daily, which is the figure stated above.

Looking at the matter in another way, it may be said that the puri- fication of a polluted water is a sort of life-insurance for the people, the value of which is equal to 10 cents per capita for each unit decrease in the typhoid fever death-rate per 100,000 which it brings about. Such a sum capitalized represents a large amount of money. In Albany, for example, where the typhoid fever death-rate has been reduced 78 per 100,000, the annual saving of life-value would be $7 . 80 per capita. Capitalized on the basis of an annual life-insurance premium of $17 per thousand, this would represent an insurance policy of about $460 per year for each inhabitant, or $2,300 for each head of a family.

Physical quality. — The figures of Table 8 show the effect of filtra-

TABLE 8.

								a	3
							oi	1	-3 •0 V 4» t;
City	Source of Supply	Type of Filter	Sample	IS 3	8 U	Odor	cU	'C 4)	HI
	Per Mil	J. Gals.
Lawrence, Mass.	Merrimack	Slow sand	Raw	10	40	3t im	S8	$11.60
	River		Filtered	0	40	2v	28	S.6o	$ 6.00
Albany, N. Y.	Hudson River	Slow sand	Raw	40	32	3v im	69	1380
			Filtered	2	24	2v	27	S.40	8.40
Yonkers, N. Y.	Sawmill Creek	Slow sand	Raw	6	30	^v im	49	9.80
			Filtered	0	3	iv	4	0.80	9.00
PouKhkeepsie,	Hudson River	Slow sand	Raw	30	S5	3v im	78	17,60
\. Y.			Filtered	0	30	iK	17	3.40	14.20
Binghamton,	Susquehanna	Mechanical	Raw	30	20	3v	S7	II .40
N. Y.	River	filter	Filtered	0	S	IV	5	1 .00	10.40
Watertown, N. Y.	Black River	Mechanical	Raw	6	70	3v	72	14.40
		filter	Filtered	0	8	IV	10	a. 00	12 .40
LitdeFalU, N.Y.	Passaic River	Mechanical	Raw	20	3S	3V	S6	11.20
		filter	Filtered	0	8	IV	10	2.00	9 20
Brooklyn. N. Y.	Baisley's Pond	Mechanical	Raw	15	31	av	52	10 40
		filter	Filtered	2	3	0	7	1 .40	9 00

76 George C. Whipple

tion on the attractiveness of waters — that is, upon the aggregate effect of their'physical characteristics :

The above figures do not pretend adequately to represent the con- ditions in any of the cities included in the list, as the analysis in each case represents only one date. They are, however, typical of what the filters in the various places are doing, and they indicate that the increased value of the water, because of its filtration, is as great as the cost of the works — in some cases it is even greater. Thus if the effect of filtration on the sanitary qualities of these waters is entirely ignored'and only its effect on their physical qualities considered, the filtration of these supplies would still be a profitable undertaking from a financial standpoint. If the sanitary qualities were also considered, the advantages of filtration would be found to be many times greater. This phase of the subject has not received the consideration it deserves, and it is this topic which the writer desires especially to emphasize in the present paper.

Water- softening. — The following figures will illustrate the financial value of water-softening plants :

Winnipeg, Manitoba —

Hardness of water before treatment 580

Hardness of water after chemical treatment and filtration 193

Reduction in hardness 387

Increased value of water due to water-softening process, per million gallons $38.70

Oberlin, Ohio —

Hardness of raw water 1 70

Hardness of raw water after chemical treatment and filtration 48

Reduction in hardness 122

Increased value of water due to water-softening per million gallons . . . $12.20

These figures refer only to water used for domestic purposes. If industrial uses also were considered the advantages of water softening would be still more evident.

At the present time there are not many water- softening plants in existence in connection with municipal supplies, but the advantages to be gained are very great, and are becoming appreciated by the managers of railroads and industrial establishments. With a better understanding of the practical benefits to be derived from the use of soft water, it may be confidently expected that during the next 10 years the number of municipal water-softening plants will very greatly increase.

The Value of Pure Water 77

SUMMARY.

In the foregoing paper attention has been called to the following propositions :

1. Pure water as compared with impure water has a real financial value to a community.

2. This value may be measured by determining what impure water costs the community.

3. There are three principal characteristics which affect the value of water to the general consumer — its sanitary quality, its general attractiveness, and its hardness.

4. A formula is suggested for computing the effect of the sanitary quality of water on its financial value to a community. It is based on the typhoid fever death-rate.

5. A formula is suggested for computing the effect of the general attractiveness of water on its value to consumers. It is based on the physical characteristics of turbidity, color, and odor.

6. A formula is suggested for computing the effect of the hardness of water on its value to the consumers. It is based on the use of soap in the household.

7. Considered from the financial aspect alone, and disregarding all humanitarian considerations, the filtration of a polluted water supply adds very greatly to the vital assets of a community ; hence, as a mere business proposition, no city can afiford to allow an impure water sup- ply to be publicly distributed.

8. The advantages to a community of having a water supply, not only safe, but also attractive in appearance, taste, and odor, are material from a financial aspect. The increased value of many waters because of the improvement in their esthetic qualities alone justifies the cost of filtration.

9. Water-softening at present does not receive the attention it deserves at the hands of municipal authorities. The economic advan- tages to be gained by removing the hardness of water are so great that, in many cases, the saving to the ordinary water-consumers justi- fies the cost of softening water.

10. The formulae here suggested and the detailed results derived from their use are not to be considered as of great accuracy, as the assumed data are not fully adequate. They are given merely to

78

George C. Whipple

show the possibihty of computing the value of pure water in terms of dollars and cents, and to illustrate the financial value of filtration and justify its cost.

TABLE 0

Depreciation Due to SA^aTARY Quality.

Values of D for Different Values of T-N in the formula D = 2.ts {T-N), Values of D in Dollars for Million Gallons.

T-N	0	I	2	3	4	5	6	7	8	9
o. .	0.00	2.75	5-50	8.25	II .00	13-75	16.50	19 25	22.00	24-75
lO. .	27.50	30.25	33 00	35-75	38.50	41-^5	44.00	46.75	49-50	52-25
20. .	5500	57-75	60.50	63-25	66.00	68.75	71-50	74-25	77-00	79-75
30..	82.50	85-25	88-00	90.75	93 50	96.25	99.00	101-75	104.50	107-25
40..	no. 00	112.75	115-50	118.25	121 .00	123-75	126.50	129.25	132.00	134-75
50..	137-50	140-25	143-00	145-75	148-50	151-25	154.00	156.75	159-50	162.25
60..	165.00	167.55	170.50	173-25	I 76 . 00	178-75	181.50	184-25	187.00	189-75
70..	192 so	195-25	198.00	200.75	203.50	216.25	209.00	211.75	214.50	217.25
80..	220.00	222.75	225.50	228.25	231.00	233-75	236.50	239-25	242.00	244-75
90. .	247-50	250.25	253-00	255-75	258.50	261.25	264.00	266.75	269.50	272.25
100. .	275.00	277-75	280.50	283-25	286.00	288.7s	291-50	294-25	297.00	299-75
110. .	302 . 50	305-25	308.00	310.75	313-50	316.25	319.00	321-75	324-50	327-25
120. .	330.00	332.75	335-50	338.25	341 -00	343-75	346 - 50	349-25	352 - 00	354-75
130. .	357-50	360.25	363 ■ 00	365-75	368.50	371-25	374-00	376.75	379-50	382.25
140. .	385.00	387-75	390-50	393 25	396.00	398 - 75	401 - 50	404 -25	407 . 00	409 - 75
150..	412.50	415-25	418.00	420-75	423-50	426.25	429.00	431 - 75	434-50	437-25

TABLE 10.

Esthetic Deficiency Due to Turbidity.

Values of P^ for Different Values of t in the Formula ^, = 5V /. Per Cent of Objecting Consumers.

Tur-
bidity	0	I	2	3	4	5	6	7	8	9
0 - . - -		5-00	7 OS	8.66	10.00	11-15	12-20	13.20	14. 10	15.00
10 ... .	15.80	16-55	17	30	18.00	18.70	19	35	20.00	20.60	21 -20	21.75
20 ... .	22.35	22.91	23	45	23 95	24-45	25	00	2545	25.95	26.45	26.90
30 ... .	27-35	27-80	28	25	28.70	29.15	29	55	30.00	30.40	30.80	30.90
40	31-60	32-00	32	40	32.75	33.15	33	50	33 90	34.25	34-60	35.00
50	35-35	35-70	36	05	36.40	36.70	37	OS	37.40	37.70	38-05	38.40
60 . . - -	38-70	39-05	39	35	39 65	40.00	40	30	40.60	40.90	41 .20	41.50
70	41-83	42-13	42	42	42.72	43 00	43	30	43.58	43.87	44.15	44.44
80 ... .	44.72	45-00	45	27	45-55	45-82	46	09	46.35	46-73	46.80	47.16
90	47-43	47.69	47	95	48.21	48.47	48	73	48.98	49-24	49-49	49.74
100 . . .	50-00

Tur- bidity	0	10	20	30	40	50	60	70	80	90
TOO . . .	50-00	52.44	54-77	57-00	59.16	61.23	63 24	65.19	67.08	68.92
200 . . .	70.71	72-45	74-16	75.82	77-45	79 05	80.62	82.15	83.66	85.16
300 . . .	86.60	88.00	89.44	90.83	92.19	93.54	94.86	96.17	97.46	98.74
400 .. .	100.00	lOI .22	102.46	103.68	104.88	106.16	107.23	108.39	109.54	110.67
500 . ..	I 1 I . 80	112. 91	114.01	115.10	116. 18	117.26	118.32	119.37	120.40	121.44
600 . . .	122.47	123-49	124.49	125. 49	126.49	127.47	128.45	129.42	130.38	131.33
700 . . .	132.27	133-22	134.16	135.09	136.01	136.93	137.84	138.74	139 64	140-53
800 - ..	141.42	142-21	143.17	144 . 04	144.91	145.77	146.62	147.47	1 48 . 1 2	149- 16
900 . . .	150.00	150-83	151-65	152. 47	153.49	154.11	154.91	155.72	156.52	157-32
1,000. .	158.11

The Value of Pure Water

79

TABLE II.

Esthetic Dkficiency Due to Color.

c Values of p^ for Different Values of c in the Formula ^^ = —

Per Cent of Objecting Consumers.

Color	0	1	2	3	4	5	6	7	8	9
o . . . .		OS	1 .0	1.5	2 .0	25	3.0	3-5	4.0	45
lO . . . .	50	5-5	6.0	6.5	70	7.5	8.0	8.5	9.0	9 5
20 ... .	10.0	10.5	II .0	11.5	12.0	12.5	130	13-5	14.0	14-5
30 —	150	15.5	16.0	16. s	17.0	17.5	18.0	18.5	19.0	19.5
40 —	20.0	20. s	21.0	21. S	22.0	22. s	23.0	23 5	24.0	24.5
50 ... .	25.0	25-5	26.0	26.5	27.0	27-5	28.0	28.5	29.0	29. 5
60 ... .	30.0	30. 5	310	31. s	32.0	32 5	33 0	33.5	34.0	34.5
70 ... .	35.0	35-5	36.0	36.5	37.0	37.5	38.0	38.5	39.0	39.5
80 ... .	40.0	40.5	41 .0	41. 5	42.0	42.5	43 0	43-5	44.0	44-5
go	45.0	45-5	46.0	46.5	47.0	47.5	48.0	48.5	49.0	49. 5
loo ...	50.0	50.5	51 0	51-5	52. 0	52.5	53 0	53 S	54.0	54-5
110 . . .	55. 0	55.5	56.0	56.5	57. 0	575	58.0