of the blood, or of their toxin by some antitoxic power, or
more generally to both. And the relapses are probably due
to decrease in both of these factors of resistance.
ii6 THE PARASITIC INVASION IN MAN [Sect.
It is not within our province to discuss these factors now.
The general experience that many cases improve without
treatment, and merely after good food and rest in bed, proves
that such factors exist ; and, as already stated, if they did not
exist every case would certainly perish.
Many diseases — tuberculosis, trypanosomiasis, relapsing
fever, leprosy, etc. — exhibit such variations in intensity. We
picture to ourselves a long struggle between the invaders and
the opposing force — first one side triumphs and then the
other, and death or recovery finally ends the contest.
Just as rest and good food encourage the resistance, so,
probably, anything which weakens the patient — fatigue, chill,
heat, dissipation, other sickness — tends to encourage the
parasites. Educated patients often declare that their fever is
brought on by such causes. Military surgeons recognise that
when infected troops are despatched upon arduous military
duties numbers of them begin to fall sick at once, even though
it may not be the season of fresh infections. Travellers and
planters complain of the same thing as regards their porters
or coolies ; and medical men notice the frequency of malarial
relapses after typhoid fever, venereal diseases, child - birth,
accidents, etc. â–
I think that external heat probably tends to encourage
relapses. Although the temperature of the body remains much
the same, the parasites may possibly be stimulated by the
heat in some way, owing to the fact that warm climates are
specially suitable to them and to their dissemination by
mosquitos. Thus in my experiments on birds in 1898, I
noticed that when the birds were taken to the cool climate
of the Himalayas their parasites greatly diminished, and when
they were brought back again to Calcutta their parasites
increased again. In the hot weather in Calcutta 11^ of the
sparrows examined contained P. danilevskyi\ but early in the
cool weather few were found to be infected. This point
requires exact experimental study. A. Caccini [1902] thinks
2o] CAUSES OF RALLIES AND RELAPSES 117
that excessive heat does not have much effect, but notes
the frequency of relapses in benign tertian in spring ; and it
seems to me that the onset of warm weather rather than the
heat itself is most likely to be the encouraging factor, since
the germicidal power is probably heightened after the first
relapses in spring. It is interesting to note that in India there
is generally a small but well-marked rise in the admission
rates of troops at the commencement of the warm weather in
March to April. For example, I take at random the Native
Troops on the North- West Frontier, Indus Valley, and North-
West Rajputana (" Reports of the Sanitary Commissioner with
the Government of India "), and give the averages for the three
years 1900- 1902.
AVERAGE ADMISSIONS.
Strength Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.
16,428 379 205 191 285 401 326 420 396 565 1,397 1,152 735
Here there is a marked increase between March and April,
that is, at the commencement of the hot dry weather, long
before the advent of the breeding season of Anophelines. It
may, however, be due to the awakening of infected hibernating
mosquitos (section 21).
Why, if the germicidal power is strong enough to destroy
large numbers of the parasites, does it not destroy all of them ?
I presume that, as in other forms of life, the hardiest individuals
often succeed in withstanding the opposing forces unless they
are too strong ; that a few of the plasmodia survive until the
germicidal power is fully developed — when the entire brood is
exterminated.
What is it that determines the length of the rally, or
apyrexial period ? Probably {a) the number of plasmodia
left alive at the end of the previous attack, and {b) the strength
of the germicidal power of the patient. If the parasites are
reduced to about the number of protospores originally
inoculated, and the germicidal power weakens rapidly, the
former will now multiply again at a rate somewhat similar
ii8 THE PARASITIC INVASION IN MAN [Sect.
to their first rate of proliferation — so that the length of the
apyrexial period should be roughly equivalent to the original
incubation period. If the germicidal power is strong, or is
fortified by quinine, the plasmodia left alive after the previous
attack should multiply so slowly that the apyrexial period may
be indefinitely prolonged. Under such circumstances a few
parasites may perhaps continue to breed for months in a
patient, without causing the smallest symptoms, and yet, if
the germicidal power is reduced for a moment, may be able
at any time to multiply again sufificiently to produce another
attack. The same thing probably happens with irregular or
insufficient quinine medication. The long intervals which fre-
quently occur can be explained equally well on this hypothesis.
We do not know why the resistance of the host should
be affected by his general state of health or by external
conditions. That such is the case affords another argument
against the idea that the relapses are due merely to some
normal development of the parasites. The study of immunity
in malaria should yield important facts for the study of
immunity in general. I merely touch upon the subject here.
Cases are often reported in which the patient during the
relapses suffers from malaise, dyspepsia, and so on, rather
than from fever, although the parasites are numerous enough
to be found. In such, I presume, the antitoxic power of the
blood has become better developed than the germicidal power.
Cases in which a relapse immediately follows a chill or accident
are explained perhaps by the hypothesis that the chill or
accident has immediately reduced the antitoxic (not necessarily
the germicidal) power of the blood.
(6). Frequency of the relapses. — All clinicians recognise that
in untreated or badly-treated cases a considerable number of
relapses may occur before complete recovery is established.
Cases giving a history of say three to ten relapses are very
common. But we are indebted to A. Caccini [1902] for
giving us a very able and full analysis of this question. In
2o] FREQUENCY OF RELAPSES 119
1899- 1901 he was entrusted by Professor Ballori, Director-
General of the Hospitals in Rome, with the care of a special
department for the study of malaria, and set himself to the
task of examining these questions methodically. The cases
studied by him were as follows : —
Type First Attacks Relapses Total
Quartan 525 330 855
Benign Tertian . . . 1,086 934 2,020
Malignant .... 2,275 i»429 3,704
Totals . . . 3,886 2,693 6,579
Many of the cases were given quinine, so that this proportion
will not accurately represent the proportion of relapses in
untreated cases — which should be much larger. The author
excluded all cases in which the parasites were not found.
He endeavoured to ascertain as exactly as possible the history
of each case, not only by enquiry of each patient, but by
reference to hospitals and localities where they had been
treated ; he kept many cases in hospital for long periods, and
he gives a careful record of his results, I will mention here
only his untreated cases.
Two cases of benign tertian, untreated, were kept under
observation in bed in hospital for nine months. The following
table gives the duration in days of their successive relapses and
rallies : —
Rel. Ral. Rel. Ral. Rel. Ral. Rel. Ral. Rel. Ral. Rel.
Case I . . . 17 13 16 16 14 19 15 17 12 10 19
Case 2 . . . 15 17 16 15 18 20 17 6 10 9
Both patients were robust, had the same diet, and were not
allowed to suffer from cold or damp or intercurrent affections.
Nineteen similar cases studied for shorter periods showed the
same phenomena. The author notes that out of 120 untreated
cases of benign tertian, the relapses occurred more frequently
in the spring (March to May). He thinks that the rallies con-
tinue from 5 to 18 days, which he calls apyrexial periods of
short interval. But he also notes and distinguishes rallies of
I20 THE PARASITIC INVASION IN MAN [Sect.
long interval, lasting 3 to 4 months, which he says occurred in
152 persons who had not revisited malarious places in the mean-
while. The rally lasted under 2 months in 60 cases ; under 70
days in 29 cases ; 90 days in 20 cases ; and from 90 to 120 days
in the remaining cases.
He also quotes the case of a band of 75 Calabrian workmen
who had become infected at Foggio in July 1900, and had sub-
sequently taken quinine until December. On 8th September
1901, 14 months after the first infection, 62 of these men were
exposed to a severe wetting with rain, while the remaining 13
found shelter. On loth September and the six following days,
every one of these 62 men was attacked with relapses. The
intervening rallies since their last attacks had lasted for 7
months in 2 of the men ; for 7-5 months in 3 ; for 8 months
in 12; for 9 months in 10; for 10 months in 15; and for 13
months in 20. Those who had not been wetted by rain escaped.
It is supposed that none of the cases had been subject to re-
infection.
In quartan the author found that out of 118 cases kept
without treatment relapses occurred only in 8 which had been
exposed to chills and other predisposing causes. He thinks
that quartan relapses chiefly as a consequence of such causes
— unlike tertian, which relapses under any conditions. The
length of the rally depends upon the care which has been taken of
the patient ; but the author quotes six cases of relapse occurring
after long intervals (6 to 10 months). He says also that in
some cases of experimental inoculation of quartan by him (no
reference) the incubation period could be prolonged at pleasure
by keeping the subject at rest in bed, etc.
Regarding malignant fever, Caccini gives the following
lengths of the rallies in 44 untreated persons : —
Days .
5
6
7
8
9
10
II
12
13
14
15
16
17
18
19
20
21
Cases .
9
I
5
I
7
I
I
I
I
I
I
I
2
6
1
3
2
The author suggests that there is some radical difference
between relapses at short and at long intervals, but I cannot
2o] FREQUENCY OF RELAPSES 121
find that he proves his case. He agrees with all observers in
attaching great importance to the determining factors of relapse,
which he says are,
(i) absence of good food and health,
(2) gastro-intestinal troubles,
(3) fatigue,
(4) traumatism,
(5) sudden chills, damp, and wetting,
(6) change of climate,
(7) certain foods and medicines (?) due to intestinal
disturbance,
(8) intercurrent sickness.
To these I would add : —
(9) anxiety, sorrow, shock or fright,
(10) alcoholic and other excess,
(11) sudden exposure to a tropical sun,
(12) premature cessation of quinine.
It is, of course, understood that these determinants act by
reducing temporarily the host's resistance, but we do not know
exactly how they do so.
Caccini endorses the common opinion that quartan and tertian,
especially the former, are most prone to relapses. In malignant,
the original illness is more severe and (probably for this reason)
relapses are not so common. I will refer to his important results
with quinine in section 23.
Celli [1901] gives a useful table showing primary cases,
relapses and pernicious cases in the Roman Hospitals for
1 892- 1 898. The following are the totals for the seven years : —
Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Total
Primary 864 676 724 793 809 696 2,912 5,157 4,440 3,875 3,074 1,967 25,987
Relapses 581 358 319 341 353 205 522 949 1,031 1,156 989 716 7,520
Pernicious 21 — — — i 45 62 67 64 56 32 330
Thus, excluding the pernicious attacks, out of a total of 33,507
cases, 75% were reported to be primary and 25% relapses — that
is, three to one.
f22 THE PARASITIC INVASION IN MAN [Sfxt.
Relapses are referred to in many other works, but the informa-
tion given is scanty. The illness being generally slighter, patients
do not often come to hospital in consequence of it ; and in most
cases quinine distorts the clinical picture (section 22).
Relapses after long intervals are frequently mentioned. The
case of P. T. Manson (section 16, case 5), who had a relapse
nine months after the original infection, followed by three
months' quinine treatment, is well known. In one case at the
Royal Southern Hospital I found the parasites in a patient who
said that he had left the tropics four years previously (records
unfortunately lost). In Britain we have many returned officials
who say that they continue to have occasional attacks after
return from malarious colonies — more than twenty years have
been mentioned to me by three persons ; but as each attack
lasts only for a day or two it is generally impossible to verify the
point microscopically. My father left India in 1880, but in 1889
still complained of occasional attacks. A little later I witnessed
one of these — a sudden severe rigor followed by high fever and a
typical sweating stage. The blood was not examined, but the
paroxysm was almost certainly a quartan or tertian one. So
far as I remember there was no determining factor in this case.
21. Average Duration of Untreated Infections.— Persons
may therefore remain infected for several, and possibly for
many years ; just as some may recover spontaneously after
the first series of attacks. But from the public health point
of view an important question remains to be answered, namely,
how long does the infection continue in untreated cases on the
average} That is, if 1,000 persons, simultaneously infected,
were to be removed simultaneously to a perfectly healthy
area, where they are kept untreated, how long would their
infections continue.? It is impossible to say at which particular
moment the last parasite dies out in a patient's blood, and
no adequate researches have been made on the point. We
are therefore forced to rely upon certain calculations.
2i] DURATION OF INFECTION 123
The cases of malaria at Ismailia from 1900-1905 are given
officially as follows : —
Years .
1900
1901
1902
1903
1904
1905
Cases .
. 2,284
1,990
1,551
214
90
37
In 1901-1902 quinine was largely distributed in the town,
but after my visit at the end of 1902 the Anophelines were
banished, the quinine also being continued to those who took
it. Since 1905, and probably since 1902, there has been no
endemic malaria, such cases as occur being found only in
persons infected outside, or being merely relapses. Hence
we infer, as a very rough calculation, that of the 1,551 persons
who were ill in 1902, less than 1/7 (214) remained ill in 1903,
and less than 1/17 (90) remained ill after two years. But here
quinine expedited the recovery on one side of the account ;
while on the other side of it many of the cases that remained
were probably due to external infection.
Another way to estimate the reduction is, I suggest, to
consider the statistics of some country where fresh infections
suddenly cease owing to the abrupt onset of a sharp winter and
the consequent cessation of mosquito breeding and mosquito
biting. I select as an example the Punjab, a large malarious
area in the extreme north-west corner of India, and give
the figures of admissions furnished by the annual " Reports
of the Sanitary Commissioner with the Government of India."
The years 1899 to 1903 have been selected because after
the latter date quinine began to be administered much more
regularly. Nevertheless, subsequent figures show very similar
monthly variations — as in fact are shown by the same statistics
for other parts of India.
The ratios under each monthly average are obtained by
dividing by it the following monthly average — so that these
ratios give the monthly factor of increase or decrease in the
number of admissions.
The striking regularity with which the maximum is reached
in October and the minimum in February will be observed.
<
l^OO "100 t^
On
1^00 fv. i-H CO
On
O vO t^vO "-
kj
Tj- OnnO -^ On
O
^ L«-> "1 ClO
»— <
<
CO CO VO ON M
ON
H
b
â– >^ iri u-i Tj- u->
l-Tt
o
1-1 •* On n ►-•
CO
h
H
"
O ON CO Tf Tj-
o
OO
o"
O OOO I- vO
r^
ON
cj
ON I^ 1-1 ON t^
ON
11
u-i -^ ro u-i ^
00
Tj-
0}
vonO VO on >-i
b
Q
" 00 ro N 00
'^
b
Q
" -^ «
'-'"
to 1^00 M OO
vO
VO
CO nO CO O OO
M
„
>"
« OO O "^ On
^
NO
^
CO ~ M CO -^
CO
NO
o
M vO \0 "
rv
b
o
CO ^ r^ -
q^
b
Z
2;
«
M M C« 1-1 N
*^
t^ "^OO \0 N
lO
ON
M t-voo CO ON
VO
VO
^
- NO CO '^ -1
CO
00
^
>- COCO VO t^
VO
NO
CJ
u-i CO t-^O On
00
b
o
- ^ CO 0;00^ "^
On
b
o
-~
o
c? 'f CO i-T pT
rT
(S t^ ON t^ O
ON
VO
a
en
vO O " O r)
o
NO
•J
CO CO O O NO
00
t-^
M O 00 O -i-
t^
^
Q,
VO rv. cooo^ O^
nO_
u^vO m\0 ^-o
lO
l-T â–ºT CP k^ >^
KH
rt On t^ OnnO
VO
NO
bb
ro i-ovO ro
OO
ON
>
bi)
O I^nO •* ri
NO
3
t^
?
t-;; o^ o^ -^ q_
-^
<
t:}- CO ltinO â– ^
"^
'"
<
"^ "-T rT "^ i-T
^
CO ^00 coco
1-4
NO
_>^
O On M ^ —
Ml
o
i33
CO On O O "
CO
CO
"5
CO -t -^ O CO
CO
â– p
On t>vCO 1-1 r^
o^
Ln CO Tf rj- -T^
1-
^
1 — >
mT hT
<
1)
P< -1 00 t-x O
n
r^
cu
Tf -I- UTO NO
t^
o
(S « o « o
VO
N
3
»- uo On r^ Tj-
Lr\
(N
1
5
r^NO VO t^ i^
00
*
-1- fi -i- ri CO
CO
C/5
«"
c
g
>-.
ii-» u-t ui O O
On
t^
>,
NO r) -^nO On
VO
CO
c^
S
00 -^ -* n U-.
NO
On
rt
:^ OnnO VO On
t^
r»
u-i M CO CO CO
CO
b
H
.^
nO â– ^ OnOO CO
NO
'-,
W
^
>
„
'^
f~^oo rv. ri Tt
^
On
's_
On VO tv.0O r^
CO
Tf
a
"-> Tj- ON -> O
NO
P
<
a.
O 00 •* N On .
ON
»-<
<
CO «-i CO n c-i
C4
<
Tt M On VO n
VO
^
nO O nO CO -
N
CO
«
On VO t~^ f< P>
VO
1^
O â–º- nO CO -i-
t^
VO
2
ON â– ^ C< 00 CS
HH
NH
f) w r) « i-i
^
CO C4 NO â– * CO
VO
"
_Q
CO nO O 00 -*
ON
CO
J2
rr M >-i O nO
„
M
4J
r^ t^ ON — CO
rs
CO
►I 00 t^ •* On
NO
t-^
fct,
*"* »-• ^
fc
VO r) Tj-vO CO
â– ^
^
•^ r) u^ TfNO
N
CO
CO 1-1 r^ p^ "*
VO
t^
rt
-* CO — On n
-^
vr\
1 — .
nO •^- ■* On t^
00
NO
1 — .
CO — rj- -1 «
c-<
b
vO vnco t^ VO
NO
b
^
O r) 1-1 \0 r^
£
T^f ~ O NO N
o
bo
OO vO CO CO On
-*
bb
rt « vooo OO
CO
4>
q^>^ -^no^ q^
ON
c
N CO O "^ VO
CO
0^
v^
OO^ncTnO^ VO r^
NO
Ul
rf r4 nO CO O
CO
^ 1
1
55
Tt ■«*■■* -"l- -^
"*
V
i~*
On Q « (S CO
ON o o o
to
_o
.
On O "I N CO
ON o o o
be
nJ
<«
rt
o
>i
CO O^ O^ On On
u
V
00 ON ON ON On
b.
V
K-
u
>
nl
>
a
<
Pi
1
<
Pi
><
<
CO
W
o
5
U
<
H
O
[-1
q^ tv. O^ -^ "^ 00
o
Q
t->.00 â– <* "^ N fO M
•*oo o "^ r^ o O
o
\0 »J^0O vO vO 00 Q
o
M \0 r» -1 O -i- 00
\0 00^ ooo o^ O^ o
mo r^ rh u-> r-^ Ln
00 r^ N CO \0 O <^
vr^ Cl^OO \0 fO l^ ■„
M
^
t^ CO O O OO vO •*
1 — .
two vO 00 t^ u-i M
O "- CO O O fO 'i-
5
vO ■-• \0 M O^ "^ '^
ro O vO \0 â–º- "- 00
M vO fO M '-' " OO
CO -* O ^ ^ i-n Q
<
00 Tf- CI t^ O M r'
CO -^ "^ ^ ^^ â– ^ M
1
r^ N O «J^ - "
Tl- ca CO •* CI fO „
J3
vO O O -' 'J^ vO C^
CT^« C» t^ M vO p
CI C< CI CO CI N „
5
u
en
ir^vO •* ■* <^ O N
O C» ro O t~^ (^ ^
CI O <> ("i <^ "-^
« LovO 00 vO O
u-i u-> •^ t^ >-< •^
CO i/i ro cP cr CO
Ov O " «!' CO 00
0^ O O O O C3 o
oo o^ O O O JiJ -^
" " " " " > rt
<: P^
126 THE PARASITIC INVASION IN MAN [Sect.
For the European troops the maximum (835) is 6-47 times
the maximum (129), which occurs four months later. For
the native troops, the maximum (2,955) is 6-41 times the
minimum (461) — almost the same. For the prisoners the
maximum (1,045) is 3*93 times the minimum (266) — a smaller
ratio.
Thus, for 1,000 admissions in each class during October
the following numbers were admitted in February: —
October February
European troops . . . 1,000 154
Native troops .... 1,000 156
Prisoners 1,000 254
A certain amount of quinine was doubtless given to all
these classes, and many of the Europeans were probably
moved to hill stations after April to May. It is easy to under-
stand the small rate of recuperation of the prisoners — a
constantly changing class drawn from the lowest ranks of
the population and depressed by their imprisonment. Hence
I infer that the rate of recovery of the European and native
troops during the four months is probably too favourable
for an estimate of the same rate among a large untreated
native population. On the other hand, the quinine' treatment
of the prisoners may be cancelled by their circumstances,
so that their rate of recovery should be something similar
to that of the general population. We may therefore, perhaps,
assume as a rough but convenient estimate for the latter that
what may be called the reduction factor is about 0*25 in four
months — that is, that 1,000 cases should be reduced to 250
in that time by spontaneous improvement.
We observe in the statistics given above that there is an
almost constant but slow increase in the admissions from
February onward. Whether this is at first due to heat relapses
or to infections by hibernating Anophelines (section 20) is
not known ; but if no such increase were to take place, we
might suppose that the same rate of recovery would hold for
22] ENLARGEMENT OF THE SPLEEN 127
the rest of that year. In other words, we might suppose that
any number of patients brought to England would be reduced
by recovery to one quarter (0*25) every four months — that
1,000 cases would become 250, 62, 15, in four, eight, twelve
months respectively, and so on. This is, I fear, a very rough
estimate, but I know of no other method of making any
estimate at all.
Hence we may perhaps infer that 3/4 of our untreated cases
will recover in four months; i5/i6of them in eight months; 31/32
of them in a year, and so on. But these figures are based upon
the gaol statistics of the Punjab — that is, upon those of natives,
most of whom have probably had malaria in childhood. What
may happen in the case of a completely non-immune and
untreated population I cannot say. For such we might adopt
a still lower rate of recovery, say 1/2 or 50% every three
months ; that is, 1/2 would recover in three months, 3/4 in six
months, 7/8 in nine months, and 15/16 in a year. On the whole,
this is probably the nearest estimate.
Celli [1901] gives the admissions into the Roman Hospitals
during thirteen of the years between 1864 and 1898 — numbering
nearly 93,000 altogether. I subjoin the monthly totals for the
reader to study in connection with the above remarks.
Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Total
4.673 3.732 3)9o6 3,921 3,575 2,553 8,844 17,678 15,203 12,755 9,531 6,621 92,992
22. Enlargement of the Spleen and Liver.— This is a con-
dition of importance in the prevention of malaria, as it often
enables us to estimate the amount of the disease in a locality.
We must now note the following points regarding it.
The average weight of the spleen of a healthy man of