mosquitos had been allowed to lay their eggs and die [1896].
The first case was, by a coincidence, attacked with fever ; but
as all the rest proved practically negative, I decided that
no reliable evidence had been obtained one way or the
other.
A. Bignami, in a paper discussing Manson's hypothesis
[1896], stated that he had failed some time previously in
causing infection by the bites of mosquitos brought from
malarious places. He believed with King that the insects
bring the poison from the marsh, and I presume that his
mosquitos were therefore collected at random. This hypothesis
was not in accordance with parasitological teaching.
The first correct experiments on mosquito-inoculation were
performed by myself in August 1896. At that time I began
to abandon part of Manson's hypothesis according to which
the insects take the parasites from man and deposit them in
water ā I began to think that the insects take the parasites
from man and also inoculate them into man. A number of
several kinds of Culex and Stegomyia were therefore allowed
to feed on various cases of malaria ā one of which contained
all the three species of parasites ā and were then fed, after being
kept for several days and on several occasions, upon Mr Appia,
Assistant - Surgeon to my hospital in Bangalore, India. The
result was entirely negative, the mosquitos being of the
17] MOSQUITO INOCULATIONS 79
wrong kinds, and the interval between the feedings too short
[30th October 1896].
The first successful mosquito inoculations were those of
birds, Calcutta, June to August 1898. Four sparrows and a
weaver-bird, which had frequently been found to contain no
parasites, and had been often used for controls in my laboratory
for that reason, were bitten towards the end of June, on several
nights in succession, by numbers of Ctilex fatigans heavily
infected with P. danilevskyi. On 9th July they were all found
to contain "swarms of the parasites." The experiment was
next repeated on many more birds under proper conditions,
with the following results : ā
(i). Out of twenty-eight healthy sparrows twenty-two or
79% were infected in this manner.
(2). One of the six sparrows that escaped the first experi-
ments was infected on a second trial (the remainder
died).
(3). Out of two crows and four weaver - birds, free from
P. danilevskyi, one of the crows and all the others
became infected.
(4). Out of five sparrows, originally containing a very few
P. danilevskyi, four showed a much more copious
infection after the experiments.
The experiments were quite decisive. At that time in
Calcutta I found these parasites only in fifteen out of one hundred
and eleven wild sparrows, and then only in small numbers in
the blood. But, as I used large numbers of heavily - infected
mosquitos, the produced infections were extremely copious, as
many as ten to sixty parasites being counted in each field of
the microscope (oil-immersion lens). The incubation period
was always four to seven days. This success was reported in
England by telegram, and was published by Manson at the
end of July 1898.
The following experiments have subsequently been per-
formed on men.
8o OBSERVATIONS AND EXPERIMENTS [Sect.
Case I. Bignami [November 1898]. ā Some months later
Bignami succeeded in infecting a man by following my methods.
A number of mosquitos caught in infected houses at Maccarese
were fed on a malaria-free inmate of the Santo Spirito Hospital,
situated in a part of Rome where there is no malaria. They
were fed on him nightly from 26th September to the end of
October 1898, and belonged to Anopheles inaculipenms, and two
species of Culicines. The subject was attacked with slight
fever on 31st October, and severe fever the next two days ;
quinine being given on 3rd November, In spite of careful search,
the parasites (malignant) were not found until 3rd November,
about forty hours after the first marked rise of temperature.
Several relapses occurred later.
Case 2. Bastianelli, Bignami and Grassi [1898- 1899]. ā
Anopheles niaculipennis caught at Maccarese were fed from
13th November to 2nd December 1908, on a healthy subject
in the same hospital. Fever commenced on 3rd December and
continued daily until 9th ; quinine on the 7th. Mild tertian
parasites found on first day of fever. Parasites found also
in the mosquitos.
Note, ā This fortunate experiment revealed that the tertian
parasites also are carried by A. niaculipennis.
Gametids appeared on 6th December.
Case 3. Ibid. ā Seven A. niaculipennis caught in an infected
house at Tre Fontane were fed, each one once, on loth, nth,
13th December 1898 on a subject in the same hospital. The
protospores were found in the salivary glands of two out of
three of these insects examined. Subject attacked with fever
on 29th December, the mild tertian parasites being found next
day ā two sets. Gametids on the 31st.
Case 4. Ibid. ā Three A. inaculipennis were fed on a case
of crescents between loth to i8th December 1898, were
incubated at 30"" C, for two days, and were then re-fed on a
17] MOSQUITO INOCULATIONS 8i
healthy subject on 2nd and 5th January 1899. All three
mosquitos were found to be infected, and two of them had
protospores in their salivary glands. Subject was attacked with
fever from 14th to i8th January. Quinine on i6th. Malignant
parasites found on 15th and disappeared under quinine on
17th, No crescents found.
Case 5. P. Manson [1900]. ā Ten A, viaculipennis were fed
on a case of double benign tertian in Rome on 17th, 20th,
23rd August 1900, and bit P. Thorburn Manson in London
on 29th and 31st August, and 2nd and 4th September. Also
thirty-five of same species were fed on a simple benign tertian
case in Rome on 6th and 7th September, and on the same
healthy subject in London on loth and 12th September. Fever
began on 13th September, and continued on 14th, 15th, i6th
and 17th. Benign tertian parasites found on i6th for the first
time. Quinine on 17th. Relapse nine months later.
Case 6. Rees [October 1900]. ā A second subject, Mr
Warren, was bitten by the second batch of mosquitos fed on
previous subject at about the same time (not given). Fever
commenced on 28th September after about fourteen days' incuba-
tion, and seemed to have continued for some time. Parasites
in phagocytes on 30th. Benign tertian parasites on 2nd October
and subsequently. Quinine on 3rd. Author says that this
case showed more parasites than the previous one, as he was
bitten by more mosquitos.
Case 7. Fearnside [1901]. ā The following series of seven
cases were obtained at Rajamundri, India, in i9(X)-i90i. Un-
fortunately, the Anophelines were not identified, but the author
states that only one species was used. Source, benign tertian,
copious, sporulating on 17th December, when the Anophelines
were fed. Insects were re-fed on the author himself on 20th
December and ist and 8th January 1901. Tertian fever began
on the 14th. Benign tertian parasites found on 18th. No
quinine.
F
82 OBSERVATIONS AND EXPERIMENTS [Sect.
Note. ā Author had suffered from mahgnant malaria with
haemoglobinuria in 1891 ; but that, apparently,
did not affect the present experiment.
Case 8. Fearnside [1901]. ā Source the same, and Anophelines
fed on same date. Subject (an Indian) bitten on 28th December.
Tertian fever on iSth January. Benign tertian parasites on
2ist.
Case 9. Ibid. ā Sources, two cases of benign tertian.
Anophelines fed 12th and 13th December. Bit subject (Indian)
on 27th. Benign tertian on nth January. Benign tertian
parasites on 13th. In notes, author says that only one
Anopheline inoculated the subject.
Case 10. Ibid. ā The same source. Anophelines infected
1 2th December and bit subject (Indian) on 26th. Tertian
fever on 20th January, and tertian parasites on 25th. Apparently
the same (one) mosquito used as in previous case.
Case II. Ibid. ā Source, double infection of malignant and
of benign tertian. Anophelines infected i6th December, and
bit subject (Indian) on 28th. Fever on 9th January, and
malignant parasites on nth; doubtful tertian parasites on 12th.
Case 12. Ibid. ā Same source. Anophelines fed on same
date and bit subject (Indian) on 28th December. Fever, 13th
January. Malignant and benign tertian parasites on 14th.
Note. ā Of two other cases experimented with by Fearnside
one appears to have failed, and one to have
contracted fever after fifteen days, the finding
of the parasites not being noted. Author states
that protospores were found in all the Anophe-
lines used.
Case 13. Buchanan [1903]. ā Experiments done at Nagpur,
India. Source, malignant. Anophelines (species not given)
fed 25th December 1901 to 8th January 1902, and bit subject
17] MOSQUITO INOCULATIONS 83
loth to 17th January. Fever 27th January, and malignant
parasites.
Case 14. Buchanan [1903]. ā Source, malignant. Anophe-
lines fed 13th to 23rd January 1902, and bit subject 24th to 27th
January. Fever 12th February, and malignant parasites.
Case 15. Ibid. ā Source, malignant. Anophelines fed 24th
January to 4th February 1902, and bit subject 7th to 19th
February. Fever 19th February, and malignant parasites.
Case 16. Ibid. ā Source, malignant. Anophelines fed 27th
January to 4th February 1902, and bit subject 4th to 20th
February. Fever 23rd February, and malignant parasites.
Case 17. Ibid. ā Source, malignant. Anophelines fed 29th
January to nth February 1902, and bit subject 12th to 23rd
February. Fever 3rd March, and malignant parasites.
Note. ā Three attempts to infect from malignant sources
failed. In two other cases, however, there was
no fever after the mosquito inoculations, but
crescents were found after twelve and eight days.
These are not accepted here.
Case 18. Ibid. ā Source, quartan. Anophelines (? species)
fed 30th January to nth February 1902 ; and bit subject from
14th to 27th February. Fever 25th, and malignant parasites
{not \.\\os,& foimd in source).
Note. ā Four other cases from quartan sources failed.
Anophelines not identified.
Case 19. Ibid. ā Source, benign tertian. Anophelines fed
8th to 20th January 1902, and bit subject 21st to 25th January.
Fever 4th February, and benign tertian parasites.
Case 20. Ibid. ā Source, benign tertian. Anophelines fed
15th to 25th January 1902, and bit subject 28th January to
I oth February. Fever 15th February, and malignant parasites
{not XhosQ found in source).
84 OBSERVATIONS AND EXPERIMENTS [Sect.
Case 2 1 . Buchanan [1903]. ā Source, benign tertian. Anophe-
lines fed ist to 13th February 1902, and bit subject 15th to 25th
February. Fever 23rd February, and benign tertian parasites.
Case 22. Ibid. ā Source, benign tertian. Anophelines fed 2nd
to 14th February 1902, and bit subject 17th to 22nd February.
Fever ist March, and benign tertian parasites.
Note. ā Five other attempts to infect from tertian sources
failed.
Case 23. Schiiffner [1902]. ā Experiments in Sumatra with
two species of Anophelines, apparently Cellia kochii Donitz,
and a Myzomyia. Source, benign tertian. Anophelines fed (?)
15th July (?) 1902, and bit subject, the author himself, on
two days (?) 25th July. Fever, double tertian, nth August
and benign tertian parasites.
Case 24. Ibid. ā Same source, and Anophelines fed same
date, and bit another subject on same date. Fever on same
date, single tertian ; and tertian parasites.
Case 25. Ibid. ā Source, malignant. Infected Anophelines
bit subject (Chinaman) on 20th, 21st August (?) 1902. Fever
5th September, and malignant parasites 7th.
Case 26. Jancso [1905]. ā A long series of admirable
experiments on the effect of temperature on the development
of the parasites in Anopheles inaculipen?iis \ carried out at
Kolozsvar, Hungary, Source, benign tertian. Six Anopheles
fed 23rd September (?) 1904, and kept at 21Ā° C. for twenty-
three days. Subject bitten i6th October, had fever on 15th day.
Benign tertian parasites on i6th day.
Case 27. Ibid. ā Source, malignant. Fifty-two Anopheles
fed 26th October (?) 1904, kept at 30Ā° C, and bit subject 3rd to
13th November. Fever 15th ; malignant parasites i6th.
Case 28. Ibid. ā Same source. Anopheles fed on 29th
17] MOSQUITO INOCULATIONS 85
October and kept at 30'' C. Subject bitten 6th to 8th November
by more than twelve. Fever i6th, and malignant parasites
on 1 8th.
Case 29. Jancso [1905]. ā Same source, and mosquitos.
Subject bitten by six on i6th November. Fever 22nd, and
malignant parasites 26th.
Case 30. Ibid. ā Source, malignant. Anopheles fed 15th
November and kept at 22-24'' C. Subject bitten by seven on
30th. After ten days, fever and malignant parasites.
Case 31. Ibid. ā Same source and mosquitos. Subject
bitten by one Anopheles on 3rd December. After fourteen
days, crescents in blood.
Case 32. Ibid. ā Same source and mosquitos. Subject
bitten on 6th December by two Anopheles. Quinine 1-2
grams on 12th and 13th. Fever and malignant parasites on
18th.
Case 33. Ibid. ā Source, malignant. Anopheles fed 24th
September at 30^ C, and afterwards kept at 20Ā° C. Subject
bitten on 21st October by one Anopheles, strongly infected.
Quinine i gram on 28th and 29th. Fever on 4th November
and crescents on 6th.
Case 34. Ibid. ā Source malignant. Anopheles fed 24th
September and ist October at 30" C, and then kept at
15-17Ā° C. Subject bitten 7th November by two of these with
glands infected. Fever and crescents on iSth.
Case T^^. Ibid. ā Same source. .^4Ā«<7//z^/^j fed 25th September
and ist October and kept at varying temperatures from 8-30' C.
Subject bitten by two of these (both infected) on 13th and iSth
October. Fever 24th and crescents 25th.
All these cases have been verified by me in the original
literature ; but there are probably some other experiments
which have been overlooked.^
' See section 65 (5).
CHAPTER IV
THE PARASITIC INVASION IN MAN
18. The Onset of the Invasion. ā It does not lie within the
province of this book to give a full description of the pathology
and symptoms of malarial fever ā a subject which is dealt with
in many works. But before proceeding to our proper theme,
we shall do well to examine a certain number of questions,
especially some upon which little stress has been laid in the
publications referred to. We should begin by attempting to
obtain a clear picture of the onset and progress of the
parasitic invasion in man.
Our first care should be to consider the number of organisms
engaged in the invasion ā a subject which has been much
neglected. For example, in the eighty-six cases of successful
experimental inoculations of men, I cannot find a single one
in which correct estimates of the number of parasites, in the
source, the carrier, or the subject, have been even attempted.
Certainly, these experiments prove the main theorem, that
the parasites cause the disease ; but they might easily have
been used to obtain many valuable pathological data in
addition. We are thus forced (at present) to rely largely
upon calculation for our figures.
(i). The imviber of parasites in the mosquito. ā The number
of parasites which enter a mosquito when it feeds on a patient
depends {a) on the amount of blood sucked up by it, and {b) on
the number of parasites in that blood. The insects feed during
variable periods ā for a few seconds if disturbed, or perhaps
86
Sect. i8] ONSET OF INVASION 87
during the whole night upon a sleeping person. While feeding
they void, every ten seconds or so, some of the fluid part of
their meal ; so that during a whole night they may possibly
consume much more blood than one stomach-full. A mosquito
fully gorged in this manner may (perhaps) consume several
cubic millimetres of blood. It would be easy to settle this
point by allowing a number of mosquitos to feed during
I, 2, 3, 4 . . . minutes, and then killing and weighing them
and their dejecta. For example, as R. Newstead has shown
("Reports Li v. Sch. Trop. Medicine, Liverpool," 1905, vol. xvii.
p. 25), the tick Ornithodorus tnoubata weighs 0*027 grams
before feeding and o'26o grams after feeding. D. Thomson
and myself find, in researches now being conducted in my
clinic in Liverpool, that 6-7000 crescents per c.mm. of blood
is not an exceptionally high number. Supposing that only
half of these are females, it is still very unlikely that so many
could ever develop, after being fertilised, in one insect ā a large
or very large proportion probably perish from phagocytosis and
other causes in the stomach contents. The largest number of
developing zygotes found by me in one insect {C. fatigans
gorged on blood with P. danilevskyi) was 445. Ten of these
mosquitos fed all night on a bird with moderate parasites
(i/iooo haematids) were found to contain an average of 29
zygotes each ; and ten of them fed on a bird with many
parasites (1/50 haematids) contained an average of 100 zygotes
each [May 1898]. But I do not know what proportion of the
parasites in these birds were sexual forms.^ The number
of zygotes found in Anophelines naturally fed on human
blood does not very often, I think, exceed fifty ā but this is
probably due to the manner of feeding. The number of
ingested sexual parasites which reach maturity and develop
protospores depends {a) on the susceptibility of the insect, {b)
on the temperature, and {c) possibly many other conditions.
Jancso [1905] finds that the zygotes develop best at 24-30Ā° C,
^ See reference to paper by S. T. Darling in section 65.
88 THE PARASITIC INVASION IN MAN [Sect.
temperatures above and below these limits retarding the process
(A. macidipennis) \ and that they die if the mosquito is kept
constantly below i6' C. after feeding. On the other hand, they
often continue to grow if the carrier is subjected merely to an
intermittent low temperature. Regarding the susceptibility of
various species of mosquitos much work remains to be done.
(2). The number of protospores in the salivary glands. ā I may
have overlooked references, but do not know that the average
number of protospores in matured zygotes has ever been
exactly estimated. I should give the number (merely as an
impression) at about a thousand. Not all of these effect an
entry into the salivary glands. Here, again, there seem to be
no exact counts. I fancy that more than 10,000 of them will
seldom be found in the glands of a single Anopheline, while
often there may perhaps be only a few hundreds at a time
(subject to correction).
(3). How many protospores enter the human blood? ā This
must depend (a) upon the number of spores in the biting
insect's salivary gland, and {b) upon the number of times it is
allowed to bite its victim. I think that mosquitos inject their
poison before commencing to suck. If this is the case, an
insect which bites a person several times (as, for instance, when
he is asleep) is likely to inoculate many more protospores than
one which succeeds in biting only once. In the former case
several thousand spores may perhaps be introduced ; in the
latter case perhaps only a few.
But not all of the spores which have entered are likely to
live. Probably many perish by falling outside the blood stream
or by becoming a prey to phagocytes.
(4). Further development of the protospores. ā F. Schaudinn
observed that if protospores are taken from a mosquito's
salivary gland and are mixed with blood under the microscope,
many of them may actually be watched entering the haematids,
where they become the young intracorpuscular parasites familiar
to students. These now begin to grow and to develop a second
i8] ONSET OF INVASION 89
generation of spores, which should be called deutevospores. The
latter attack fresh haematids, within which they grow, and
develop a third generation of spores, which we may call
tritospores ; and so on indefinitely.
(5 ). The number of spores produced by each species of parasite. ā
The parasites of malaria are described in detail in the text-
books. Since the time of Golgi, all observers admit that
they belong to three types at least, each of which differs
morphologically, and markedly so, from the others. I assume,
for the sufficient reasons frequently given, that these three
types are three different species. The number of spores pro-
duced by each species is variable, and different authors give
different figures (probably the number has never been accurately
estimated). I adopt the following names and figures for the
present : ā
Plasmodium malariae Laveran, 1881. The Quartan Parasite ;
produces say 6-12 spores every three days.
Plasmodium vivax Grassi and Feletti, 1890. The Benign
Tertian Parasite ; produces say 15-20 spores every two days.
Plasmodium falciparum Welch, 1897. The Malignant
Parasite ; produces say 6-20 or more spores every two days.
Some authors consider that there are two if not three
varieties (or ? species) of malignant parasites. I am inclined
to agree with them, but have not yet satisfied myself
sufficiently on the point to admit it in my classification.
(6). The onset of the invasion. ā Let us suppose that a
mosquito has inoculated several thousand protospores, but that
only one thousand of these have succeeded in entering the
haematids. After two or three days, according to the species
of parasite, each will produce a variable number of deutero-
spores. But, probably, not all of these will succeed in infecting
fresh haematids ; many may be devoured by phagocytes, or be
destroyed by other agencies, while passing from one haematid to
another. And the same thing is likely to happen with every
successive generation of spores. Hence only a proportion of
90 THE PARASITIC INVASION IN MAN [Sect.
the spores actually produced are likely to enter fresh
corpuscles.
Suppose that i,ooo protospores of P. vivax have entered
haematids, and that, on the average, only lo out of the 15-20
spores actually produced in the successive generations succeed
in entering fresh haematids. The parasites should then multiply
as follows : ā
No. of days 024 6
No. of parasites 1,000 10,000 100,000 1,000,000
No. of days 8 10 12
No. of parasites 10,000,000 100,000,000 1,000,000,000
and so on.
In the case of P. malariae the multiplication should be
slower, if the number of spores given above is correct. But
in that of P. falciparum the data are too uncertain for
calculation.
(7). The number of haematids in an average man. ā According
to accepted standards we have ā
{a) I c.mm. of blood contains 5,000,000 haematids.
{b) The specific gravity of blood (male) is about I057"5.
(<:) The total amount of blood weighs about 4-9/100 of the
total weight of the body.^
From these data we calculate that 3,000,000 c.mm. of blood
weigh 3" 1 7 kilograms, and will be contained in a man
weighing 6474 kilograms (142 lbs. English, or about 10 stone),
and that these 3,000,000 c.mm will contain 15,000,000,000,000
haematids.
This may then be taken as the normal number of haematids
in an average healthy man of about 64 kilograms, or 10 stone,
in weight. The reader should endeavour to form some tangible
idea of this enormous number. If he were to try to count it at
the rate of 100 a minute day and night without cessation, he
would have to spend 285,000 years over the task !
^ The ratio of blood to body weight was formerly estimated at about 1/13, but the
estimate has recently been reduced in consequence of better methods of measurement.
i8] FEVER-PRODUCING NUMBER 91
(8). The lowest number of parasites required to produce the
first illness. ā We can scarcely imagine that one protospore, or
even many thousands of these minute bodies, can produce any
marked symptoms in the patient ; and, in fact, we learn from
the fundamental experiments that as a rule no symptoms
appear for some days after the moment of inoculation ā this
period being commonly called the incubation period. But as
the parasites increase at every generation, a time must come
when the number of them will suffice to cause illness. Can we
determine this number? No serious attempt has yet been
made to do so, or, indeed, to find any exact correlation between
the number of parasites and the amount of sickness. There
are many difficulties in the way. It is not always easy to
estimate the number of parasites, and still less to know when
precisely the illness commences. In most of the experimental
inoculations, some illness, or even sharp fever, has occurred
before any parasites have been found by the microscope. On
the other hand, Elting has been able to find the parasites in