The corresponding ratios, namely, 4'5/2"o, 8/5, 17/9 (taking
o P. as the normal temperature), are not dissimilar. The pyro-
genous limit here suggested is lower than my arbitrary figure
of 150,000,000; but I may have underestimated the infection
of Mendez, and the weights of the subjects are not recorded.
I04 THE PARASITIC INVASION IN MAN [Sect.
The corresponding paroxyzm of Mendez himself rose to
40'i' C. and lasted five hours ; but his was an old case, and
cannot be compared with the others.
We do not know the nature of the toxin, but I have always
thought it likely to be some soluble constituent of the plasmodin.
It is generally supposed that the various species of plasmodia
produce toxins of different strength (or virulence) ; but we
should await better proof of this.
(5). Antitoxins. ā Just as the body certainly produces some
germicidal property capable of limiting the number of the
parasites, so it must certainly produce some antitoxic
property capable of destroying, or at least eliminating, their
poison. We do not know its nature, but infer for the follow-
ing reasons that its power increases with time : ā
(i) From the time of the ancients it has been observed
that malarial fever begins with remittent fever, due
to the overlapping of the paroxysms, and ends with
intermittent fever, due to the shortening and separa-
tion of the paroxysms. This shortening of each
attack of fever is not always accompanied by a
decrease in the number of parasites, and must
therefore be due to something which reduces their
effect on the body.
(2) Old cases, especially children, often scarcely feel
the paroxysm, although considerable fever may be
present.
I have observed in many untreated cases that the number
of parasites continues to increase, although each attack of fever
tends to become more and more mild. This suggests that the
antitoxic power of the blood tends to be developed sooner than
the germicidal power.
20. The Decline of the Invasion.ā (i). The period of regular
paroxysms. ā It is not easy to find untreated cases for study,
but from such as I have been able to examine or to read about.
2o] DECLINE OF THE INVASION 105
I infer (subject to correction by more exact numerical methods)
that the following course of events occurs : ā
After reaching a figure of say 50 or 500 per c.mm. the parasites
tend to increase more slowly owing to the increasing germicidal
power of the blood. That is, while each sporid still produces
(?) the same number of spores, a greater and greater number of
the young parasites are killed ; so that the invasion tends to
reach a limit of something like 1,000 to 10,000 mature sporids
at every generation. But this limit may not be reached perhaps
for some weeks.
In the meantime the increasing antitoxic power of the blood
tends to shorten each paroxysm more and more ; the fever, at
first remittent, breaks up into a series of intermittent attacks.
Thus, though at each paroxysm the total number of parasites
may be slightly increased as that number approaches the highest
limit, yet the parallel increase of the antitoxic power annuls
the effect of the parasitic increase, so that the patient now
often suffers from a long series of paroxysms of almost equal
intensity, which give the typical classical picture of a malarial
fever.
This period may last, I think, for some weeks, but of course
the process may be subject to modification. Thus, food and
rest may help the patient, while in some cases death may occur
in spite of quinine from an unlimited propagation of the parasites.
Fortunately, however, the period ends abruptly in most cases
even without quinine. The paroxysms become very mild, and
then, suddenly, the sporids undergo an immense fall in numbers ;
and this stage of the disease ends.
The collateral incidents and symptoms are minutely described
in pathological books, and it is necessary here to refer only to
some important points.
According to the elementary law of Golgi, a patient may
contain, not only one set of parasites sporulating every second,
or every third day (as the case may be), but two or three sets
sporulating on different days. He may also contain sets of
io6 THE PARASITIC INVASION IN MAN [Sect.
parasites of different species. The rule generally accepted is
that each set of parasites continues its own evolution independently
of other sets which may be present. But much more precise
work requires to be done on this point, and on the
following : ā
{a) The exact increase of severity, if any, of a paroxysm
due to the simultaneous sporulation of two sets of
parasites of different species on the same day.
{b) Does one set of parasites affect the body-reactions
against another set of the same or of a different
species ?
if) Do different sets of the same species tend to coalesce ?
{d) Explanation of the (apparent) extinction of one set
long before that of another.
The presence of different sets may obviously be ascribed to
inoculation of the patient on different dates. This matter has
already been touched upon, but not exhaustively, in some of the
inoculation experiments, especially in those of Di Mattei, Elting
and Jancso.
Besides fever, the parasites produce anaemia. This has often
been measured absolutely, but never exactly in correlation with
the number of parasites. It is supposed that the toxin of the
Plasmodia destroys many of the haematids which are not
mechanically broken up by the parasites themselves. More
exact researches are also required regarding the absolute and
relative leiicocytic variations in correlation with the number of
Plasmodia.
The different species produce paroxysms of somewhat
different type and intensity. Many authors quote figures, but
without correlation with the number of parasites. Thayer and
Hewetson [1895] give i r8 hours as the average duration of the
single benign tertian paroxysm, about ten hours as that of
quartan paroxysms, and twenty to twenty-one hours as that
of the malignant paroxysms. The massing of the malignant
parasites in the inner organs, and many other details, are
2o] ONSET OF SEXUAL FORMS 107
pathological matters. The enlargement of the spleen and liver
will be dealt with in section 22.
(2). The appearance of sexual forms. ā With the quartan and
benign tertian parasites, these forms begin to be seen very
early after the plasmodia become numerous enough to be
found at all. It is very doubtful whether they are or are not
produced in broods at any given moment during the course
of the invasion or during the development of a single genera-
tion ; and still more doubtful how long a single gametid can
live in the blood of the patient. The opinions of many
writers may be quoted, but, unfortunately, they are only
opinions. Schaudinn's case will be referred to presently.
The malignant gametids are much more easy to study. They
begin to appear about a week after the onset of the fever and
may remain circulating in the blood for several weeks. We
do not know what determines their appearance, but the asexual
forms and the fever frequently vanish when they appear (which
may often be due to the fact that quinine influences the sporids
more than it affects the crescents). In untreated cases, how-
ever, we may often find both forms together for many days,
though the sporids are apt to be scanty. In such cases there
may be slight oscillations of temperature, which some writers
have attributed to the crescents. Often, also, the sporids may
be still present in large numbers though they are too few to
be detected. In infected sailors in Liverpool we often find
crescents with great certainty, unless the earlier infection is
cut short with quinine. But many writers complain that in
certain localities they cannot find them as frequently as might
be expected, and I have had the same experience in India and
West Africa. To explain such cases I have surmised that the
production of sexual forms may be largely influenced by season
ā that they may be produced abundantly for a few weeks, but
only at that time of year when the local carrying Anophe-
lines are most abundant. Caccini [1902] says that several
observers have noted the absence of crescents when the new
io8 THE PARASITIC INVASION IN MAN [Sect.
malaria season is imminent, and that he himself has never
found crescents in Italy from April to June. For another
example, I found very few crescents at the foot of the
Darjeeling mountains in 1898 at a place where they had
been frequently seen by another observer a few months earlier.
But this is a mere conjecture, and one which requires long
study for verification ; and it scarcely explains why sailors
in Liverpool have crescents so regularly. The early use of
quinine ā to which sailors are not addicted ā may also explain
the paucity of crescents in many cases. Lastly, it is possible
that the sexual forms are produced only during the earlier
stages of an infection ā that in an old infection, when, so to
speak, parasites become worn out, they no longer produce
sexual forms, just as the oldest broods of human cells fail to
produce them. In my experience crescents are common during
the first few months of an infection, but comparatively rare at
later stages. Failure to find " flagellated forms " (spermatophoria)
must be generally due to faulty technics.
(3). The period of rallies and relapses. ā We have seen then
that the period of regularly repeated paroxysms generally ends
after a few weeks, even without treatment, in a great decline
in the number of sporids. The gametids, especially the
crescents, may still remain numerous, but the asexual forms
which cause the illness diminish so much that they fall below
the pyrogenous limit and cannot easily be found.
The patient now improves in health. His paroxysms
cease, his haematids increase, his spleen diminishes, and he
begins to gain flesh again. I call this the rally.
From this point two things may happen. The patient may
recover completely, or ā without reinfection ā he may suddenly
suffer from a relapse.
In the relapse all the old symptoms of the period of
regular paroxysms, sometimes less in intensity and sometimes
worse, recur. The fever may recommence with the remittent
form, and may then, as before, become intermittent. The
2o] RALLIES AND RELAPSES 109
anaemia and the splenomegaly ā not yet completely recovered
from since the first attack ā will again increase, and to a greater
degree than at first. Large numbers of sporids will again be
found in the blood, and may be followed as before by crops
of gametids. Finally, in most cases, even when untreated,
recovery, accompanied by a great decrease of sporids, will
occur again.
This is followed by another rally, which in its turn 7nay be
followed by another relapse ā and so on, indefinitely, for months
or perhaps years.
Such, in my opinion, is the normal course of malaria, whether
untreated or badly treated. But as the former kind of case
can scarcely ever be observed, my opinion is based, not upon
continuous observation of many cases, but upon correlation of
different periods in different cases. I think, however, that most
students of tropical malaria share that opinion.
During the whole of this period the emaciation, anaemia,
splenomegaly and secondary symptoms tend to increase with
each relapse and to diminish with each rally. But the increase
seems at first to be generally greater than the decrease, though
by a diminishing increment. Thus, after some months the
patient tends to reach a condition in which these symptoms
arrive at something like a fixed limit ā considerable emaciation,
anaemia, splenomegaly, oedema, dyspepsia, etc. This condi-
tion is well known as that of chronic malaria, and is only too
frequently seen in malarious places, especially among children.
In most cases, especially in children, after the fixed limit
has been reached a general improvement sets in. After all,
the disease is essentially a benign one. Each relapse now
affects the patient less and less, and the secondary symptoms
begin to decrease more with each rally than they increase with
each relapse. After an unknown average duration, complete
recovery certainly occurs in the large majority of cases, without
any serious treatment. Such cases are said to have become
partially immune.
no THE PARASITIC INVASION IN MAN [Sect.
On the other hand, deatJi occurs in many as the result of
" pernicious paroxysms," intercurrent affections such as dysentery
and pneumonia, or general weakness assisted by poverty caused
by inability to work.
Of course this picture ā or rather sketch ā is modified by
many conditions. Patients who have already become partially
immune during childhood or after many attacks, suffer much
less on reinfection ; and good food and change of climate
benefit the case, while complications such as dysentery, sprue
and ankylostomiasis, have the opposite effect. The most usual
modification is that caused by inadequate treatment, and I
continue to see many such cases in Liverpool. The history
is always the same. The patient has been instructed to take
quinine during his attack, and for a week or two afterwards.
This is insufficient, and the inevitable relapse occurs. But
the patient seldom falls into the extreme condition of wholly
untreated cases. The drug has not extirpated the infection,
but it has checked each relapse as it occurs, and has controlled
the anaemia and splenomegaly. Such especially is the condition
of the infected sailors of whom we see so many in Liverpool.
They still have relapses, but the spleen is not markedly enlarged
and the anaemia not very pronounced.
Works written on malaria in temperate and often more
civilised countries do not always give a complete picture of
the disease as it occurs among poor natives in the tropics.
This is due to the fact that the more general use of quinine
in the former tends to abbreviate the course of the infection.
On the other hand, the great majority of cases in the countless
malarious villages of the tropics are cases of chronic malaria
ā untreated, subject to frequent reinfections, and with anaemia,
enlargement of the spleen and numerous relapses. In temperate
climates, our cases are mostly in the early period of regular
paroxysms ; in the tropics they are mostly in the long-continued
period of rallies and relapses.
(4). The parasites during the rallies. ā The general observa-
2o] THE PARASITES DURING THE RALLIES iii
tion that the degree of fever depends ceteris paribus upon the
number of parasites suggests that when there is no fever the
parasites are likely to be at least scanty. Hence, as is to be
expected, in the rallies between relapses few sporids, and indeed
very often none at all, can be discovered in the small quantities
of blood examined under the microscope ā though, of course,
the gametids (which do not seem to produce fever) may still
abound. Nevertheless, when the relapse occurs, the sporids
again appear in detectable or large numbers. What happens
to them when they disappear and reappear in this manner ?
We can easily observe that the number of parasites, when
they are numerous enough to be found, often varies largely.
It is therefore equally easy to infer that when we can no longer
find them, this is due simply to the fact that they have become
too scarce to be found. But many writers seem to think that
when they cannot find sporids none exist, and have therefore
sought other explanations of the disappearance and reappear-
ance referred to.
Years ago Marchiafava and Bignami suggested that when
they disappear for long periods the parasites may become
encysted somewhere in the inner organs ; or that they may
die out altogether, and that the relapse may be caused by
spores which escape from phagocytes in which they have been
living in a dormant condition. No evidence has been given
for these views ; yet it should be easy to obtain in the malaria
of birds, which remain infected for months. In 1898 I examined
many birds in vain for the supposed encysted forms.
Golgi, Mannaberg and others thought that crescents keep
the infection alive during the rallies. Thus Mannaberg said
[1894, p. 302]: "These relapses in localities free from malaria
are to be explained only by the persistence of parasites
possessing certain powers of resistance (probably the crescentic
bodies) within certain tissues." But why are any special powers
of resistance required ; and how explain relapses with the
parasites which do not produce crescents ? Many authors.
112 THE PARASITIC INVASION IN MAN [Sect.
working, I believe, with unstained films, have described and
figured sporulation of crescents ; but others attributed this
merely to post-mortem vacuolisation. We have never found
sporulating crescents in stained films in Liverpool, though
numbers of these bodies are examined in class. B, Grassi
[1901] suggested that the relapses are caused by parthenogenetic
reproduction in the gametids.
This idea was followed up by F. Schaudinn [1903]. A
patient who had suffered long from P. vivax was attacked on
29th April and ist May, and was found to contain both sporids
and gametids. A rally now occurred, during which daily
examination disclosed varying numbers of gametids only. On
25th May these parasites were more plentiful. Next day curious
changes were noted in the female gametids, suggesting that they
were producing spores similar to those ordinarily produced by
the asexual sporids. The author considered this to have
occurred independently of fertilisation by the male gametids,
and to be due to parthenogenesis. The same day (26th May)
the patient had a slight rise of temperature to 38*4Ā° C, and in
the evening ordinary young sporids were found. Next day only
these forms occurred. On 28th May there was a typical attack
with temperature reaching 4075Ā° C, and with the usual sporids.
Next day only gametids were again found. The author care-
fully described the parthenogenetic forms and traced the
corresponding changes in the nucleus.
These observations were hailed as an important discovery by
many writers, especially zoologists. They have been partially
repeated by Maurer [1902], Bluml and Merz [1908], Harrison
[1909] and others ; and many of the younger and more confident
workers have spent much time in attempts to verify them, or
have seen objects like those described by Schaudinn. But the
objects twice seen by Maurer are not in any way proved to be
parthenogenetic forms ; while those observed in six preparations
by Bluml and Merz are supposed by the authors to be more
probably cases of production of gametids by gametids ā another
2o] THE PARASITES DURING THE RALLIES 113
hypothesis ; and these authors do not agree with Maurer.
Harrison admits that the supposod " gametoschizonts " seen
by him may be merely double infections of haematids by a
gametid and a sporid. Craig [igo6] thinks that " intracorpuscular
conjugation " may maintain the infection.
It is easy to see under the microscope objects which suggest
this or that hypothesis, but only the most patient and long-
continued labours suffice to prove the truth. Schaudinn's case
seems to me of doubtful value. I note especially that between
1st and 25th May the number of gametids varied, and was
increased on the latter date. But this variation suggests that
they were being produced all the time. The gametids are
supposed to be produced from ordinary spores ; so that we
are forced to infer that a number of sporids, some of them
generating gametids, were present in the patient's body, although
they were too few to be detected in the small quantities of blood
examined by the author. On 25th May they probably increased
in number sufficiently to induce a slight attack of fever, and
were then mistaken for parthenogenetic gametids. On 28th
May they produced a typical attack ; and that is all. The
supposed nuclear changes were reported on evidence of no great
value. The cells were not actually observed undergoing the
development which the author describes. He merely inferred
the existence of the development from a study of different cells
in what he thought were different stages of that development.
Worst of all, no numerical estimates are given. The same
author enunciated many other hypotheses on similar evidence.
I doubt whether parthenogenesis occurs with the parasites of
malaria, for the following reasons : ā
{a) If it occurs in one species it ought to occur in all. It
ought therefore to be easily observed in birds'
malaria, and in crescents. I have been examining
the latter for fifteen years and have never seen in
them anything suggesting parthenogenesis. Yet they
possess a definite and uniform outline which should
H
ii4 THE PARASITIC INVASION IN MAN [Sect.
be much modified by such cellular changes as
Schaudinn described. A. Carducci [1905], though
favourably inclined towards the parthenogenetic
speculation, failed after careful search in finding
any evidence of it in crescents.^
{b) I have frequently seen relapses occur in cases of P.
falciparum in which no crescents at all could be
found, Carducci notes the same (see section 65).
{c) In four experiments (see end of section 14) blood
containing numerous crescents only was injected
into healthy persons by capable observers without
producing any infection whatever. Yet I showed
in 1 895- 1 896 that crescents live on ice and for hours
under vaseline. Why then, if Schaudinn's hypothesis
is true, did they not infect any of these four persons ?
{d) If relapses are caused merely by some natural develop-
ment in the parasites we should expect them to
occur more or less independently of the state of
health of the host ā which does not seem to be the
case (see subsection 6).
{e) If it occurs at all, parthenogenesis ought to be common
enough to be demonstrable with ease and certainty.
This subject is of great importance in connection with the
prevention of malaria. Crescents withstand much medication
with quinine, and if they keep the infection alive in patients, it
follows that quinine loses much of its value as a public pro-
phylactic against malaria. But, while I am quite ready to accept
any evidence for it which may be offered, at present I do not
think that the Grassi - Schaudinn speculation is even nearly
proved.
The fact that the parasites disappear at each rally and re-
appear at each relapse is explained quite easily and sufficiently.
In the rallies they become too few to cause fever or to be
^ H. M. Neeb [1909] in a paper just to hand shows careful, but I think still quite
inconclusive, work in favour of parthenogenesis in crescents.
2o] THE PARASITES DURING THE RALLIES 115
detectable in the small quantities of blood examined ; in the
relapses they become numerous again, I cannot see that a
simpler or more satisfactory explanation is required. We
observe the same process in relapsing fever, tuberculosis,
trypanosomiasis and other diseases.
A good writer once said : " It is hardly conceivable that it
(the parasite) should remain in the general circulation, passing
through its ordinary cycle of existence, without causing any
symptoms whatever. Further, the failure of repeated examina-
tions of the blood of patients who have previously suffered
from malaria to reveal the presence of the parasite renders
this most unlikely." It is impossible to accept such a train
of thought. If 150,000,000 sporids cause but little fever in a
man of medium weight, 1,000 or 100 should cause none at all,
and yet will be quite sufficient to keep the infection alive in
him. But if the parasites are so few, what chance has the
medical man of finding a single one? If even 1,000 parasites
are present in a man of 10 stone weight, the chances are that
we must search 15,000,000,000 haematids before we find one
parasite. At the rate of 10,000 haematids a minute we shall
have to search for 1,500,000 minutes, or for twelve hours a
day for more than five years, before we succeed. Or we may
put it in this way ā that 150,000 examinations, each of ten
minutes' duration, must be made.
In seven autopsies on cases of " latent malaria," C. F. Craig
[1909] found only the ordinary parasites in the spleen. Four
of these contained malignant parasites, yet no crescents at all
were found in the spleen.
(5). Probable cause of the rallies and relapses. ā According to
this simple hypothesis (which I will continue to adopt until a
better one is established) the rallies are probably due to a
great destruction of the parasites by some germicidal power