T
M+ZIVDICAL jO1RXAL]
IPIDEMIC DYSENTERY IN THE FIJI ISLANDS.
If we m-alke a mistake we suffer for it, anldl can blame no
oeii but ourselves. But wlhy slhould we be called upon to
suffer for the fads and miiistakes of people at lhome ? We
lhold that furthei' teimporizing, in tlle light of our present
kniiowledge, is notliing slhort of criminiial folly.
'Witlh reference to Dr. Neave's suggestioni that tlle aamue
-%itlhin a certain area slhould be imiipounded aud destroved,
I may say that this suggestionl was miiade fully two years
ago. We asked, whben the first epidemic brolke otut in
Nyasaland, to lhave tlle game destroyed. This was refuLsed,
on1 the grounds, amolng otlher reasonis, tlhat the outbreak
'Was distinctly localized, anid that aniy interferelnce witlh
tI c game miglht cause the aniimals to sprea(l out alnd carry
lisease into clean districts. Oni the strenotlh of this
statement I stuggyested to the Actinia Goverlnor that lhe
slhould ilmlpound all the aiuimals withlin the area mentioned,
anid once and for all stamip out tlle disease by exterminatilng the animals. To this proposal I (ot a replv that while
I [is Excellency appreciated muy good intentions he did not
regard the proposal as practicable.
In conclusion, I slhould like to express my appreciation
of and gratitude for the work of tlle Liverpool Commission,
w%hich has settled once and for all many hiitlherto lhotly
dlisputed points, and also muy gratitude to tlle BrrTIsH
MEDICAL JOURNAL for the ptublicity it is givinia to the work
of Dr. Varrington Yorke.
A STUDY OF EPIDEMIIC DYSENTERY IN THE
FIJI ISLANDS,
WITH SPECIAL. REFERENCE TO ITS EPIDEMIOLOGY AND
TREATMENT.
BY P. H. BAHR, M.LA., MI.B., D.T.3M. AND H.C.At'nB.,
The Bacilli.
By plating out tlle miiucus dilutedl in sterile brotlh oni
Conradi-DrigaLski mediuim (litmus-lactose-nutrose-crystalviolet agar), dysenterybacilli were isolated from the stools
in, 35 cases and twice from the intestinal iiiucosa 1)ost
m1ortIe7n. The bacilli were recognized by their morphology,
by their reactionis with 1 per cent. solution of various
sugaars in litlmius peptone water, by tlleir agglutination
reactions with high dilutiolns (1 in 50) of a polyvalent antidysenteric serum (Lister Institute), and fturtlher by their
toxicity to guinea-pigs on intraperitoneal injection.
The dysentery bacilli isolated mainly conformed to tlle
well-known types, the one most frequently encounterec
being the Shiga-Kruse bacillus, but a bacillus identical
witlh the Hiss and Russel Y bacillus was also almost as
common. In addition to these, otlher bacilli, morplhologically and cultuLrally silmiilar, but giving atypical reactions with soltutions of certain sugars suclh as saccliarose
and maltose, wlhile agglutinating with the same polyvalent
immtune serumn, were also encountered, and in these
reactions agreeing with Strong's2 original Plhilippine
straini.
In the varieties of bacilli enlcountered the resuilts, in tllo
main, are in agreemlent with those of Shiga,3 Wllo noxW.
recognizes five main types of dysentery bacilli as differentiated by their sugar reactions.
On three occasio-ns bacilli giving identical reactions to
those cited above, but in addition exlhibiting a feeble
lactose fermenting power, were encountered. Ohno,4
during his extensive study of 74 strains of the dysenitery
bacillus, also appears to lhave encoulntered this variety,
to whclih he assigns 3 out of his 15 types.
Table qfJ?eactionis of Ma(iil Types of Dysentery Bacilli in Litmuts
Peptone Wfater conitainiing I per Cent. of Sngars afyter
tI'w entymton Houtr s' Licubation at 3, C.
Shiga-Kruse
Type.
LONDON SCHOOL OF TROPICAL MEDICINE.
THE observations recorded in this paper were made in
the Fiji Islands durinlg the year 1910. This group of
islands, situated between lat. 17-19 degrees S. and lona.
177-179 degrees E., was ceded to Great Britain in 1874,
and now constitutes a Crown Colony.
According to Dr. B. G.- Corney, I.S.O., a great autlhority
on all matters appertailling to the Pacific, there is no evidence that dysentery was originally introduced by tlhe early
European settlers,0 but it is known tihat since the early
days epidemics of dysentery have been of anniiual occurrence,
and lhave levied a heavy toll of life, niot only amona the
European settlers and the imported inidentured Indiarn
and Solomon Island labourers, but also among, the Fijian
natives themselves, and have constituted tlle chief scourge
of the colony; so lhealthy in other respects.
In former days these dysentery epidemics were of great
virulence; the newly imiiported Solomon Island labourers
were specially attacked; thuLs Daniels'1reported one such
epidemiic in which the deatli-rate was 48 per cent.
At the present day, probably owing to better mnedical
service and hygienic conditions, the case mnortality from
dvsentery of all kinds is not nearly so hiiah as formnerly,
and amongst the Fijians, accorlding to the official records,
is about 16 per cent.
During the year 1910 somue 170 well-marked dyselntery
cases were studied in detail at the Colonial Hospital, Suva,
the capital of the group; of these only 11 were found to
be of amoebic origin-that is to say, vegetative amoebae
bearing the characteristics of the Amoeba histolytica of
Sclhaudinn were found in the stools. In the dysenteric
discharges of the other 159 cases no amloebae were ever
found.
Tlle epidemuic dysentery affected miiembers of the
various nationalities-Indian, Fijian, Solomon Islanders,
Europeans, and lhalf-castes-to an equal degree as far as
could be ascertained.
In Fiji, as in mnore temperate zones, epidemic dysentery
is an institutional disease; some 53 per cent. of the patients
iunider observation were prisoners undergoing selntences in
either of the prisons of Suva.
* The Fiji Islands were discovered in 1643 by Tasmian, who, however, did not laud there. The, first Euiropeans who had intercourse
with the natives were the mutineers of the Bouniity, who had crossed
over froin Talhiti in 1791.
VIJEB. .7, 11r4
Mannite
Lactose
Sacelharose
...
-
!Hiss and Rissel's nOlo
Type Y.
TYple AT.
+
+
+
-
+
Maltose
..
=
Dulcite ..._
+
+
+
Dextrose
... Acid at first, re- Acid at first, re- Acid at first, ieMilk ,,,
turlned to alka- turned to alka- turned to alk;aline seconld day line second day line fonrtli day
The produietion of indol by these bacilli was very variable. In the
trule Shiga-Kruse tyl)e none was produced as a rule, e%-en after
lengthy incubation.
After long stubculture over a period of nine miontlhs, amI
after the sugar reactions lhad been frequently tested,
certain minor variationis of certain strains were noted,
especially witlh regard to the fermiientationi of mualtoses
and sacchlarose; nevertlheless, the reactions of the nonacid or Slliga-IKruse and the acid or Flexner type of
bacillus to dextrose and mnannite respectively renmained
constant throughout.
Clinical Observations.
Cases from whiclh the dysentery bacilli were isolated
varied greatly in their clinical miianifestations -frlont
a mild catarrlial conditioni, in wllich tlle stools Iverc
solid and coated with a small amount of blood and
mucus, to a fulminating one, in whiclh tlle call to stool
was incessant and in which the patient, overwhelmiied
with pain, phvsical exhaustion, and toxic absorption,
would succumb in from two to three days froml thIe
comminencement of the attack. The temiiperature also
variecd within wide limiits; cases were encounitered in
which it was subnormal tlhrouglhout, otlhers in wllicll it
reached 103° F. and over. No special dysentery bacilltus
was found to be associated with any clinical type of case.
No hepatic symnptoms in life or suppurative lesions of
thlis organ post moi-tcnr were ever encountered in thlis
inlfection. Only OneO case of bacillary dysentery of
a relapsing nlature w-as encountered amlongsattIme whIole
serZies. The serauml of 74 per cent. of thle bacillary cases
agglutinated- in a dliluitionl as hligh as 1 inl 200 in some
EPIDEMIC DYSENTERttY IN
FEB. 7, 1914]
THE FIJI ISLANDS.
t
r
Tax BRTISni
MEDICAL J
29R5AL
t
inistances-one
or
otlher of the dysentery organisms.
A positive agglutination did not take place until six day-s
after the commencement of symptons.
The eleven amoebic cases, on the other hand, "iave a
previous Ihistory of
a muchl more subacute and relapsing
character. Symptoms of acute hepatitis were present in
four instances, and in another multiple miliary abscesses
of amoebie origin were found post mortem.
In muakina a diagnosis of amoebic dysentery from the
mIlicrOscopic examination of the stools alone care m-ust be
exercised so as not to mistake certain round refractile
cells, probably representing disintegrated intestinal epitllelium and leucocytes, and in which redl bloodl cells are
often enclosed; these in some instances certainly bear a
close resemblance in size and form to the pathogetic
amoeba, and are at present in considerable ntumbers in
the stools of bacillary cases as well as in those of the
more clhronic amoebic disease.
Elpidemiology.
studying the epidemics of dysentery wlhiell
In
have
occCttied in Suva in recent years one is strtiuck by the fact
that the maximum incidence of the disease- oceus clurina
the months of December, January, February, Marclh, and
April, the period of the highest niean temperature and of
the largest rainfall.
In searchling for an explanation, of this fact I was led
to inLquire into. the possible sources of the bacillary
infection.
From what we know of the bacteriology of acute
dysentery it is right to assume tllat as the bacillus
escapes from the body in tlle stools these, as in the case
of typlhoid, act as the primary source of infectioln, and,
further, tllat tlle alimentary route is probably tlle one
by wllich a fresh infection takes place. The water supply
of the town of Suva is beyond reproaclh, tlherefore any
suspicioni of its contam-ination by infected faecal matter
can be dismissed.
There is, in additioni, nio evidence that the direct infection of foodstuffs plays any part in the spread of the
disease, since acute dlysentery occutrs amongst all iaces
in Fiji, thouglh their respective dietaries are enltirely
dissimilar. The Fijians for the most part live on cooke(d
farinaceouLs foods, such as the yani anid the dalo (Calocasia
eseulcdnta),- whilst tlh Indiian subsists, as elsewlhere, on
boiled rice and pulse; the Europeans on a mixed diet,
in which freslh imieat plays an important part.
In considering tlle indirect methods bv whichl infection
could be conveyed to foodstuffs suspicion fell on the lhousefly. Any visitor to the islands will bear m-e otut -when
I say that the lhouse-flies (Museca domnestica) constitute a
plague of consider-able importance. They swarm in the
'buslh and on the roadside, covering the t-aveller from head
to foot. In thle- lhou:ses every article of food mnust be protected f-orom their attentions, and it is a comnmon
rence dutring mneals to place a fly-paper on the table in
orl(er to drawv away the flies ifom tlle mi-eal. I nee(l
lhardly add that in, tlle ratlher insanitary kitchens tlley are
occur-
specially abuLiidant.
The abundance of flies in the cultivated dlistricts can
probably be explailned by the fact. tiat numerous favourable breeding places are afforded by lheaps of decomposing
suglar cane, and i: lthe unicultivated clistricts by the faccal
accumulations -whichl generally m- arlk tlhe proximity of
native villages.
There can be little doubt that in Stiva, at any rate, there
a definite seasonal variation in tlhb number of hotuseflies, and that this season corresponds with the conditions
m1ost favotirable for their propagation, th-at is the .seasonl
of highest rainfall and atmosplheric temperattre. anid as I
hlave previouLsly pointed out, it is just duiring these ns-ontls
tlat the maximum nu1mber of cases of acuite dysentery
is
occur.
Investigations on the bacteriological flora of flies capa dysentery ward of the Colonial Hospital, Stiva,
were ulndertakeln. A number of attempts were maade but
on two occasions only dlid I succeed in isolating the
typical Shiga-Kruse b)acillus fromi the lower intestinial
tiact of the fly.:
The technique adopted was based on the work of
tuired in
A
summary of
tllese observations
has already been
publiLshed iul
oJ 2'rmoip-ul
SR'upplement No. 2 of tile Jou?mata of t1e l,El(lono Sch0ool
Jle7diciiLe. NAitherby and Co., London. 1912.
C%
Graliam-l- Smith; the suspected flies were caught in batch
5
es
of four in the vicinity of a patient from wlihose stools the
Shliga
Krtse
bacilltus
had
already been isolated.
Trle
wings and legs of the flv were first remov-ed. the bodyv,
sterilized as mtlch as possible by passing it rapidly thirouh
a flame, was then placed on a sterilized slide together wit]l
a drlop of sterile saline solution.
The abdomlinal seginents;
were next separated by means of sterile needles, thits exposing the stomach and intestinal tract; a portion of the
latter was then removed on a platinum rodl, enuluified in
sterile broth, and plated on Conradi-Drigalsli mliediumi.
The Shiga-Kruise bacillus was recognized by tlle tests
already given.
In view of tlhe incom-iplete data existing at present con
necting "wild flies with the spread of disease, this resuilt
may be considered of some value wlhen taken in conijuntction witlh the seasonal distribution of tlle disease in
Fiji and tlle concomitant prevalence of the house-fly.
A niumber of Gramn-negative non-motile bacilli ere also
isolated from the intestines of these flies. These bacilli
do not ferment the sugars with gas production, and, as has
been pointed otut by Graham-Smitlh, are probably normal
inhabitants of the intestinal tract of the fly. Tlhe,
however, are ituportant in that they are liable to be mistaken for pathogenic organism-s, especially the dysentery
-
"
bacillus.
In his interestinig book, Flies anzdl Di8ease,6 GralhamlSmith has dletailed this subject at great lengthl. These
bacilli whiclh do not liquefy gelatin are divided into lactose
and non-lactose fermentinjg groups. Of tlhese, the lattet
grouip, from my point of view, are the most imupo-tant.
Graham-SSmitlh found non-lactose fermenting bacilli in
32 per cent. of flies fromi
he Three Maloin
foutnd in
l
1 ..
2
--
-
3
diarrhloea-infected houses.
Voarieties of Non-lactose Fersnettiuq Bacilli
the Intestines of iieCs cautlht in, Fiji.
A.
A.
A.
A.\., -
A..
!
A.
CAcid; alkaline fourtli to
A.
A.
~~~~~fifthl day .
Alli0line.
Acid; alkaline tlfiir
fiftlh
to
Tliese prol)y1)ly correspond to GSroup C, Suibgroul) b, Groil) B. Subj, and11l G;oup C(, SuIb)groui) ai, respectively, in Grahaml-Sm;lith s
groul)
classification.
Tllese bacilli
were
fuLrtler differentiated from the true
dysentery organism by their failure to agglutinate witlh
a
polyvalent antidyseenteric seruimil.
Of tlle bacilli fermienting, lactose witlhout gas prodtiction
anid forming perm-anient acidi'ty and occasionally clot in
milk a great niumber werc fotund. To these bacilli I nleed'
not refer at any lengtlh, as tlley are of little scientific
interest in this cotiiexion. All were tested witlh a
polyvalelnt antidysesiteric seruLmil, but witlh a negative
restult.
Late- experiments on the trausmission of tlle dysentery
bacilltus through the initestinal tract of the house-flv were
undertaken at the London Selool of Tropical Medicine.
They were conducted in a glass house in wlhichl the
temperature alpproaclled that of the tropics, reaclhing as
higlh as 91° F. in time daytinmc. Before experiments w%ere
commenced the holuse -as tborojiglly waslied out witlh
lysol, and all stray flies and otlher inse&ts were rigidly
exclhided. Thle. flies were bred andt hatehed in captivity
under as ste-ile coniditions as possible.- The pupae .vere
wvashed in dlilute m-ierculry peerchloridesolution (1 in 1.OOOn)
and placed in a sterile cage till the flies cemiergedl. Thiey
were then transferred iii batches of twenty in sterile test
tuLbes to sterile bottles fitted witlh a gauge sleeve in the
manner recommendedl by Graham-Smitlh, aud fed witlh a.ii
emiiulsion of sterile bread and water introduced into the
bottles by meanis of a sterile spoon. The flies were,
infectedl by means of bread soaked in a brotlh culture of
tIme dysentery organism. Over forty experiments wetue
other
conducted in this manner, but for some reason
not all were successful.
Two strains of bacilli belongina to tIme non-acid and the
acid or Flexner group were uised for infecting the fliestllat is, thle true Sliiga-Kruse bacillus and the Y bacillus-or
296
TiiF -oInTISSH
EPIDEMIC DYSENTERIY IN THE FIJI iSLANDS.
bacilli wlichl after frequent testing and subculture over
ppcricd of ninie miontlhs lhad given constanit fernmentationl reactions witlh the stugars. Tlhe infected mnaterial,
after being, left in conitact witlh the flies for twenityfour hiours, was removed anid sterile food was given as
l)efore.
At varyinga periods batclies of flies were dissected, anid
the conitents of the lowver inltestine were plated out in the
mannier described above, -with the result that neitlher of
these oorallisnms could be recovered from the flies' intestinies after the fiftlh day. The bacilli were recognized by
the uistial tests.,
I was uinable to obtain* any evidence of the multiilication- of these orgyanisms in the fly.
Somnewhat sim--ilar restults have been obtainecd by
(lralhan- Smiitlh in'lhis experiiiients witli typlhoid bacillus
anjd B. entcritidlis.
Naturally in these experimEnts the possibility of tlle
flies beino infected witlh tlhe dysentery bacillus in tlhe
larval state caninot be excluded; but againist this probability
aIre tlle experimertts of Bacot and Gralhami-Smnith, wlho
fouind that tlle nion-sporing saproplhytic orgauisms onlly
rarely survived in the intestines tllrouglhout the larval anid
Jupal staae, and that cultivated strainS of B. typhossts and
B. enitet-itidis fed to larvac do not strvive. Lately Tebbutt,8
inl a series of experiml-ents in whiclh larvae were fed with
cultures of B. diysenteriae, found that the organislm could
only be recovered from the pupae and imagines in a smuall
number of cases, but concluded that the possibility of
infectioni of flies by these organismiis in their breedinig
gtounds may be conisidlered a very remote oIne.
The dlysentery bacilluis tlhus recovered in the mliajority
of instances gave, wlheni tested on solutions of the sugars,
the samie ireaction as the original bacillus witlh wlichl the
ilics wvere inifected; tlerc were, lhowever, several exceptionls. A maltose fermentin)a strainv
was obtained from -a
Rly originally fed with the Shiga-Kruse bacillus. On
iftrther subculture, over a period of fifteen days, this
was lost.
Froi-m flies infected,1 witlh tlle Y bacillus a bacillus -was
recovered witlh suchl feebie power of fermienting imaanniite
that acid Nwas ouly prodltuced in this sugar after four days'
inicubationi. This proper ty it retainied for two miionthls
after isolation, but suibsequenltly lost. Otlher bacilli
capable of feormenting saeclarose in addition to miiainnite
and dextrose were also obtained; these, how-aver, are niot
of sutcl great initerest.
A Flexuer baeillu.s exhlibitinga but a poor reaction to
imianniite has been fotunid by Slhiga in tlle stools of aetite
dlysentery; aceording to the ptublislhed descriptions, tlis
baci!lns gi'-es a similarireaction to tlat I have jtst recorded.
Shiga considers that this bacillus occupies a place halfway betwecii the acid and noni-acid grouips.
That a bacillus can acquite these claracteristics by a
short sojoUlrn in the intestinal tract of the fly certainily
suggests that under certain coniditions dysenterly bacilli of
one group can acquiire the claracteristics of thle otlher.
It may be that tlis is ani instance of an adaptation of tlle
bacillus to its surroundings conmparable to tlle muutationi of
the B. tlyp7hosots wlhen grown for a lonig period on lactose-outaining m-ledia, accordinig to the researches of Penfold9
an,d Twort.10 My results are certainly suggestive of somle
su;chl process, btut it muust be nioted that the acquired
c-lharacter of this bacillus was not retained for miianiy
,,nerations; it is tlherefore necessary that these expel-iinents shlould be repeated on a mIlore extensive scale.
property
TretIntent of Actite D,sei tc)r,.
In a disease charaeterizced by clinical form-lls of suchl
vtirying severity, especially in niative patielnts belonginig to
races differing widely in plhysique and disease-resistinig
powYers, it is diffictult to compare tlle results of various
foriims of treatment. In every series, hlowever selected,
thjere will necessarily be a number of muild cases in whiclh
thje constitutional distturbaniecs are so sliglht as to cause
the minimunm amount of incolnvenienice, anid wlichl, if
nttreated, would recover in a few days. The only reliable
ethlod.woukl seem to be to compare the results of treatmen,t by diferent methods of a palAlel series of onsecuttive cases all approximlately of equal severtity.
Thle first 53 consecutive case.s, of wthichl 41 per eent.
hJad marked constitutional .symlptomls, -wvere treatedl by. thle
rountinle treatment adopted inl thec Colonial Hospital, Suvra.
[FEB
, 1914
This treatimienit consisted of the admilnistration of sodiuma
sulphate in draclim doses given every ho1Ur for the first
twelnty-four hours, and subsequently eveiy fouLr hours.
The case mnortality in tllis series was 13.2 per cent.
Tlle next series consisted of 106 cases, of whlicli 42 per
cenit. lhad marked constitutional symptomis. Somle 34
of these were treated by salines comnbined witlh cyllini in
gelatin capsules (palatinoids) in doses of twenty to tlhirty
in the twenlty-foir holiurs (1ti 60 to 90 pure cyllin); the
remaining 72 cases received in addition to this intravenous
injectionis of a polyvalent antidysenteric seruLm (kiindly
supplied by the Lister lnstitute). A dose of 20 c.Cm. was
giveen to adtults, 10 c.cml. or less to cliildren.
In apparenitly lhopeless cases injectioni of 50 to 70 c.cm.
of this serumi-in tlhe first twenty-four hours after adniissioi was followed by remarkableimprovemient. After sucl
injections no deatlhs occurred in a series of five cases,
in whliomn tlle disease was of tllh severest type, as evidenced
by tlle passage of gangrenous stools and the toxic condition
of the patient.
In tljis series of 106 cases tllere were only two deatlsa mnortality rate of 1.8 per cent. 'Neither of these deatlhs
could be ascribed to the dysenteric lesions-on'e, a Fijian
child, succumbed to an intercurrent attaclk of broncllopneumonia; the other, an Indian woman lheavily infected
witlh ankylostomes, died in uraemic convulsionis. Attention has been drawn to the occurrenlce of deatlh fronm
after m-assive doses of a polyvalent aintidysenlteric
neplhr4is
serum by Savage in hiis El Tor series.
In this inistalnce niot only was the average stay in lhospital
of cases of miioderate severity considerably, slhortened, but it
was also found that the stools resumed their. normiial faecal
consistency in a much slhorter average space of timiie-tlhat
is, after five days, as compared witlh eiglht oln the sodium
sulphate treatmiient alone. These favourable results lhave
beeis confirmlied by Willmore"l and Savage working witlh a
hell mlore extended series of cases. The case mortality
rate of dysentery in Eaypt seems to lhave been abnornmally
hiialg (70 per cent.), but under antiserum treatme'nt in 1912
to 1913 it was reduced to 12 per cent.
The oral administration of salines anid initestinal anitiseptics, togetlher with the intraveniouis injectionis of anitiserumii, appears to me to rest on a rational basis whlen the
following facts are considered..
The lesions of acute dysentery are conifinied to tlhe
large intestine, and from it the dysenteric toxins arc
absorbed; it is probably the absorption of tllese toxins
w-hich is responsible for the collapse so often encountered
in tlhis affection, especially in clhildreln. To couinteract
the deleterious effects of this toxin, anitiserumiI- is injected.
The administration of sodiuiii sulplhate is a necessary
measure to cleanise the lowver bowvel of faecal iimatter and
tlhereby to accelerate thle repair of the ulcerated miiucolus
membranie. The administration of cyllin, on tlle other
lhanid, in suclh large doses must certainly tend to inlhibit
the muitiltiplication of tIme dysentery bacilli and otlher
organismns in tlle intostinal contents.
It must be bornie in mnind.that a very acute dysentery of
primiiary amoebic origin., associated witlh the passage of
gangrenous stools, in wlich amnoebae, though present in
thle intestinal lesions p)ost mortemn, canlnot be found in tlhe
stools in aniy niuinbers durinig life, is also sometimiies met
wvitlh. In IFiji I cncounltered one suclh case. Tlherefore,
in alny giveni case of an apparently desperate nature
anid in which there is lno time to make a diagnosis by
cultuiral miietlhods, it is advisable to combine the specific
treatmenlt of both tllh amioebic and bacillary forms.
If tlie case be not of primary amoebic origin, the hypodermic injections of emet ne, if inefficacious, is completely
lharim-less.
'T'lhc comiibined emetinie and antiscrumii treatment in any
case of doubtful o-igin is therefore the imost rationial to
adopt in the circumstances.
REFE:REN CES.
I Daniiels, Pramctiiioner, vol. xlv, 1890, Pi). 343-346. 2 StroDg anid
Musgrave, Jourit, Amer. Med. A-soc.. 1900, Xxxv, pp. 498-501. Sbiga,
Phitl. -Joitrn. Sci. I, No.5, 1906, p. 425; Zeit. fir- Hg. I(. Iinfect, Ix.
1908, I, Pp. 75 and 120. Ohno, Phil. Jour n. Sci., I, No. 9, 1906, P. 951.
5 Grabamii-Smitl, Local Govereinment Board Reports (Newv
Series), xl,
1910, 1, P. 1; and xliii, 1911, IV, p. 31. 6Grahaii-Smnith, 1-'lies and(I
University .Press, 1913. 7 Bacot, Pa r1asit eley.w
DVisRase,
Camlbridge
vo 1,
iv NO 1, 1911, P. 68. d Teh)hutt, Journ1. Hyg., vo0 . Xii, 1913, lPp.
516-526. 9 Penfold, ,Toum-. Byg., vol. ii, No.-l1,l911,pl. 30. 10-'PmT
Proc. ASoc. Latd., vol. lxxix, 1907, p. 329; CenltralbIl. fiU 8akt. xl, 19)7,
p.- 508. l1 Wilimlore and Sav-age, BaITISU MEDICAL JOURNAL, 1913, I1,
