Researches on malaria Nobel Lecture, December, 12, 1902

to me at that time.

    R E S E A R C H E S   O N   M A L A R I A

41

days’ residence in the mosquito - to become more visible, and, if they were to

pass into the water as we thought they would do, to take a definite form of

resistance which ought to be easily recognizable. My new method then was to

give up the attempt to follow the newly escaped and almost invisible motile

filaments, and to dissect the mosquitoes, not at once, but after the lapse of

some days, during which time the motile filaments should by hypothesis de-

velop into something more tangible. I proposed then to feed the mosquitoes

as before on cases with crescents in the blood; to keep them alive for some

days; and then search their tissues for any parasites which might occur in them.

The parasites found, it would be easy to determine whether or not they are

derived from the motile filaments, simply by ascertaining whether or not they

also occur in mosquitoes of the same kind fed on healthy blood. Throughout

the investigation it was of course necessary to employ only what I called in

bacteriological parlance "sterile mosquitoes", that is mosquitoes freshly hatch-

ed from the larvae in captivity and therefore not contaminated by previous

feedings.

Such was the procedure now adopted; but the difficulties involved even in

it were very great. As the situation of the sought-for parasites could not be

indicated with any certainty, it became necessary to search for them through

all the tissues of each insect examined - to scrutinize by a powerful micro-

scope, one by one, all the minute cells composing the huge aggregate of which

the insect consists.* To investigate a single insect thoroughly in this manner

required at least two hours’ exhausting and blinding work. Added to this dif-

ficulty, I had no clue as to the form and appearance of the object which I was

seeking for; nor was I even sure that the kind of insect under examination was

amenable to the infection at all. I was looking for a thing of which I did not

know the appearance in a medium which I did not know contained it. In short

it was a mere blind groping for some clue which I trusted fortune would give

in the end. As an instance of the difficulty of such work I may mention that

neither the organisms of yellow fever, which is now known to exist in a par-

ticular kind of mosquito, nor the Pyrosoma of Texas cattle-fever, which is

known to exist in a tick, have yet been found in these animals, though long

searched for by competent observers. Nevertheless, I am confident that, hope-

less as the method may appear, it was the only one capable of solving this

difficult problem.

At the outset of the investigation it was necessary for me to become thor-

oughly acquainted with the normal histology of the mosquito - for which I

* Under a magnification of a thousand diameters a mosquito appears as large as a horse.

42

    1 9 0 2   R .R OS S

had again to trust to my own observation; in spite of all efforts no literature

on the subject could be obtained by me. It was also necessary to note and study

the ordinary parasites of these insects - of which I found a number during the

ensuing years. Indeed, at the commencement of the work I found one which

required careful working out. It was a pseudo-navicella occurring in the mal-

pighian tubes of the brindled mosquitoes (Stegomyia). After a little study it

was ascertained that pseudo-navicellae have no connection with the parasites

of malaria, being the sporocysts of a species of gregarine. Next year Manson

published an account of these interesting organisms taken from my letters to

him

26

. I refer to them also in my publication at the end of the year

24

.

The second method alluded to above was based on the following considera-

tions. According to Manson’s more advanced hypothesis, the motile fila-

ments, after living some days in the mosquito, would probably pass from it

into the water, on the surface of which we then supposed it usually died after

laying its eggs. Such water then ought to be infective to human beings, either

when ingested, or perhaps when inhaled as a vapour. It would be easy to test

this speculation by experiment. I caused a number of mosquitoes, both of the

brindled and grey varieties, to feed on a selected patient, and then kept them

in large jars containing water at the bottom, until they died one by one. The

water was then exposed to sunlight and otherwise allowed to remain in the

condition of marsh water. Different batches of fed mosquitoes were intro-

duced into the jars from time to time so as to make sure that the water should

indeed contain the parasites which by hypothesis should escape from the in-

sects. In May 1895 I gave draughts of this water to three natives who volun-

teered themselves for the experiment. All of them declared that they had not

suffered from fever for years.* Strangely enough one of the men developed a

mild but marked attack of fever in eleven days, the parasite being found in his

blood. I was naturally much pleased with the success of the experiment and

began to hope that the mode of infection had been found; but the failure of

many subsequent attempts of the same kind forced me later to reject any defi-

nite conclusion on the point (section 11).**

11. Bangalore (1895-1897). Progress of

 work. Possessing abundance of material

* The experiment was justifiable owing to the slight degree of illness usually produced

by malarial fever in natives when properly treated.

** Owing to the interest of Surgeon Major Owen, the Maharajah of Patiala had at this

time offered the government of the Punjab to employ me at his own expense to study ma-

laria in his dominions. The government of the Punjab, however, refused the offer.

    R E S E A R C H E S   O N   M A L A R I A

43

together with plenty of leisure, I was now progressing excellently with my

researches (though without definite results) when I received my first interrup-

tion. On the 9th September 1895, I was placed on special duty by the Govern-

ment of India to combat a serious outbreak of cholera in the large town of

Bangalore, and also to report on the general condition of sanitation there. This

duty was of great interest and value to me because it afforded me an unrivalled

opportunity for examining in the closest detail the general sanitation of a

tropical city - an experience which has stood me in good stead during later

years. For four months, however, I was so busy with my new labours that I

had little time for research. Bangalore I already knew well, having indeed

made some of my first studies of malaria there when staff surgeon of the town

from 1890-1893. I now easily ascertained by the light of Laveran’s great dis-

covery that the cases of fever which I had attributed to intestinal auto-intoxi-

cation were nothing but examples of aestivo-autumnal infection among par-

tially immunized natives. I found also that, as at Secunderabad, my brindled

and grey mosquitoes abounded all over the place. I dissected a few mosquitoes

as time allowed; and when my more arduous sanitary duties began to dim-

inish in March 1896, I found that I could give an hour or two a day to the

work. My results, however, remained constantly negative, in spite of the

closest scrutiny of many mosquitoes. At the same time I continued my at-

tempts to produce infection by water.

In March 1896, Manson delivered the three Goulstonian Lectures at the

Royal College of Physicians in London, and again put the case of his hypoth-

esis in an admirable manner, supporting his arguments largely upon my ob-

servations of the previous year

2 6

. He wrote to me frequently for fed mos-

quitoes, which I sent to him whenever I could. He also urged me to keep on

the track of the flagellated spores; advised me to try infection experiments

with the insufflation of dried and powdered mosquitoes, and with the vapour

of an "artificial marsh" in which fed mosquitoes had died. These devices did

not appear as promising to me as they evidently did to him. It was scarcely

likely that dead mosquitoes could do much in regard to the dissemination of

malaria in nature, at least in the form of dust, owing principally to the fact

that dead insects seldom escape the ants in the tropics. All dust, moreover, is

generally subject to the intense heat of the sun which, except in the presence of

water, must be very inimical to most unprotected organisms. The small

amount of time at my disposal was therefore devoted to the methods already

attempted.

Towards the middle of the year I had made nineteen experiments with a

44

   1 9 0 2  R .R O S S

view to carrying infection by drinking water; and together with three more

cases, I described these in a publication at the end of the year

30

. Water, in

which mosquitoes fed on cases of malaria had died, or which contained large

numbers of the pseudo-navicellae of the gregarines of mosquitoes, was given

to various persons by the mouth. The majority of the attempts were entirely

negative: but nevertheless a slight but noticeable reaction did occur in three

of the whole number of twenty-two cases. This still remains a very curious

circumstance; but the facts were published exactly as they were found, with-

out the influence of the "personal equation". At the end of the paper I sum-

marized my results and decided that the positive reactions, though interesting,

were too few and too slight to warrant any definite conclusion.* I am now

inclined to think that they may have been due to the following circumstances.

The persons on whom the experiments were made were generally low-caste

indians who required a fee before drinking the water and also an assurance

that they would receive more if taken ill. Now it is well recognized that many

natives are constantly infected with malaria and get relapses on any extra-

ordinary demand being made upon their systems, as by fatigue, chill, or dissi-

pation. I have even heard it stated by medical men possessing large experience

of natives that they can often produce fever in themselves by exposure when

they wish to do so. In this case at all events it was possible that some of the

subjects spent their preliminary fees in dissipation, thus producing the sup-

posed reaction after the experiments.

These results not being as decisive as I had expected from the first experi-

ment of the kind made in the previous year, I began to consider whether some

other route of infection was not possible or probable; and it soon grew upon

me that Mansons’s induction was exigent only as regards the entry of the

parasites into mosquitoes, and that his secondary hypothesis regarding their

escape from the insects and their infection of man through drinking water

was not so strong. I quickly thought of several other routes for infection -

which will be examined presently; and first I considered it possible that the

insects, previously infected from diseased persons or possibly from other mos-

quitoes, might then inoculate the parasites into healthy persons during punc-

ture, or might deposit them on the skin during haustellation. It was easy to

*My actual words were, "While we cannot dream of stating definitely on the strength

of these experiments that there is something connected with the mosquito which is capa-

ble of imparting fever, the three positive results are still curious and tend to be in favour of

the truth of Manson’s theory." Yet one of my Italian critics has attempted to prove, by

ignoring this passage, that I pretended to have established infection by drinking water.

    R E S E A R C H E S   O N   M A L A R I A

45

test this view immediately by experiment; and early in August I made a small

series of observations which were published in the same paper

30

.

A number of mosquitoes all bred from larvae in captivity, and of all the

kinds which I could collect (many specimens of brindled and grey mosqui-

toes) were fed upon several patients with numerous parasites in their blood.

One of these patients had all three kinds of parasites in him; and I specially

employed this case, as well as many varieties of mosquitoes, in order to in-

crease the chances of one at least of the species of mosquitoes present being

appropriate for one at least of the species of parasites. After feeding, the insects

were kept alive for one or two days and were then applied in considerable

numbers on two occasions to Mr. Appia, Assistant Surgeon of the Bowring

Civil Hospital at Bangalore, who courageously volunteered for the experi-

ment. Mr. Appia had suffered from malarial fever some years previously, but

not since then; so that if he should be attacked by fever shortly after the ex-

periment, it would be strong evidence, if not proof, in favour of the inocula-

tion theory. He remained, however, absolutely free from fever. He was then

bitten by five mosquitoes which had been partially fed immediately before on a

case of crescents - on the supposition that the insects may carry the infection

mechanically, as the tsetse fly carries nagana; but the result was again negative.

Lastly two other individuals were bitten by mosquitoes fed from three to five

days previously; still without effect. I judged then, either that infection is not

produced in this way, or that the proper species of mosquitoes had not been

employed, or that they had not been kept for the proper period after feeding;

and I proposed to return to the subject again. It should be noted that these

experiments of mine were made quite independently, and before I had heard

of the theories of King and Bignami - as indeed was stated in another publica-

tion of mine at the end the year

32

.

In July 1896 Bignami’s criticism of Manson’s hypothesis, referred to in sec-

tion 6, appeared in Italy

29

. I heard nothing about it whatever, until I received

Manson’s letter of the 12th October, which was accompanied by a translation

of the critique. Bignami’s paper was not a profound one, and consisted only of

a copious and dexterous rendering of ideas which were new only to those who

had not already fully considered the subject. His objection to Manson’s theory

was based principally on Grassi’s loose speculation that the motile filaments

are the result of the death of the parasites in vitro. As this was a vital point in

the chain of reasoning I now set to work to examine the subject experiment-

ally, and was soon able to show that the escape of the filaments depends on

certain proper conditions, and not at all on the death of the parasites. Thus

46

    1 9 0 2   R . R O SS

they escape more readily when the specific gravity of the blood is altered,

either by the abstraction of water by partial evaporation or, as Marshall prov-

ed, by the addition of a little water. On the other hand they do not escape at

all, even when the parent cells perish, so long as the blood is kept scrupulously

unchanged. In order to prove this, I drew the blood from the finger into a

small mass of Vaseline placed upon the skin, and then mounted the whole for

the microscope in such a manner as to prevent the blood coming even into

momentary contact with the air. The result was that not a single crescent

emitted motile filaments or even underwent the preliminary change of spher-

utilization, although it was evident they all died after a time.* This experiment

completely disposed of the death-agony theory of the Italians. Previously to

this, however, Sacharoff had shown that contrary to Grassi’s statements, the

filaments do contain chromatin; but I could not procure a copy of his work

32

.

I should add that after long observation of the filaments I could never bring

myself to believe that they are merely the result of the spasmodic movements

of dying protoplasm; and this tale was in fact never anything but a gratuitous

assumption.

These researches were published later

32,33

, and were confirmed by Manson

and Rees in London

37

.

The conditions required by the crescents for emitting filaments were now

clearly seen to be those obtaining in the mosquito’s stomach, where the blood

is rapidly altered by abstraction of water; and I therefore continued my work

without further reference to Bignami’s objection.

His view that infection may be caused by inoculation had already been con-

sidered and experimented on by me, as just mentioned. But it should be under-

stood that Bignami’s hypothesis (which was the same as that given long pre-

viously and much more strongly by King) was very different from mine.

King and Bignami thought that mosquitoes bring the poison from marshes to

man; this speculation had not occurred to me until I read Bignami’s paper in

October, and then it did not appeal to me at all, because it was self-evident

that the connection between malaria and marshes could be sufficiently ex-

plained by the fact that mosquitoes breed in stagnant water. My speculation

was that mosquitoes become infected from men (according to Manson’s in-

* If the preparation was opened and the blood momentarily exposed to the air within

some hours after abstraction from the patient, the crescents could be seen at once to re-

sume their functions. But if this experiment was delayed about 24 hours, the crescents no

longer reacted, and indeed showed clear evidence of death by their vacuolization and

other structural changes.

    R E S E A R C H E S   O N   M A L A R I A

47

duction) and possibly also from other mosquitoes, and then communicate the

parasites to healthy persons - perhaps by inoculation. It will be seen which

view is right; but in consequence of my negative experiments, the inoculation

theory was not much favoured by me until I made my researches in the Sigur

Ghat (section 12).

My duties at Bangalore continued for a year and a half At first placed upon

special duty to report on the sanitation of the town (80,000 inhabitants), I was

afterwards appointed officiating Residency Surgeon there and was required

to reorganize the whole of the sanitary arrangements, to create a health de-

partment, to participate in a committee designated to reconstruct the munici-

pal regulations, and to contend against several outbreaks of cholera. Conse-

quently I did not possess as much time as at Secunderabad for my researches

on malaria, but nevertheless in addition to the experiments last referred to,

I was able to dissect many hundreds of mosquitoes in pursuance of my princi-

pal plan of campaign. Several agents were employed to collect the larvae of as

many kinds of mosquitoes as possible, especially from several spots whence

most of the cases of fever came; and these insects, belonging to many species

of the brindled and grey groups of mosquitoes,* were all tested by direct

feeding on cases of malaria, especially aestivo-autunmal. But though each

insect was examined with the utmost care, almost every cell being scrupulous-

ly searched for parasites, the results still remained entirely negative.

Towards the end of my stay in Bangalore, as failure followed failure, I was

naturally forced to reconsider the whole basis of my work. But no; the most

critical examination of Manson’s induction failed to exhibit any flaw in the

fundamental reasoning. The gametocytes, and the process by which the mo-

tile filaments escape from them after the blood is drawn from the patient,

could only be meant for infection of the mosquito. There was no other ex-

planation. Nature does not create these complex phenomena for nothing; and

the theory must be - was - sound. What then was the cause of my repeated

failures? Was it possible that the kinds of mosquitoes which I had tested hith-

erto - very many kinds - were all of the wrong species?

The reasons for and against this view were as follows. In all the districts and

towns of India in which I had served or stayed during fifteen years - Madras,

Bangalore, Moulmein, the Andamans, Secunderabad, Upper Burma, Bom-

bay, Poona, Calcutta, Karachi, Quetta, the Nilgherry Hills, malaria was un-

doubtedly present, especially among the natives ; and in all of them without

* Not once as yet had I come across the dappled-winged mosquitoes ; though, be it noted,

many of my larvae were collected from ditches and the edges of ponds.

48

   1 9 0 2   R.R OS S

exception I remember to have noticed mosquitoes belonging to both the grey

and the brindled classes. This naturally suggested a connection between the

disease and the insects; but, on the other hand, were not the later perhaps too

common? So far as I could ascertain, the disease was generally limited to cer-

tain spots and localities (by no means always near marshes); whereas the in-

sects were everywhere, and were indeed often commonest at points were ma-

laria was rare, as in the houses of Europeans. After all may not the true malaria-

bearing variety or varieties have been overlooked by me? Possibly they were

comparatively rare species, or species occurring only at a certain season - a

hypothesis favoured by the well-known fact of the seasonal variation of ma-

laria. Now, as I was fully aware at the time, malarial fever is a relapsing disease

in which attacks continue to occur for years after infection; so that it does not

follow by any means that the infective variety of mosquito must always be

present in a locality, even though numerous cases of malarial fever are present.

And it was to be specially noted that most of the cases occurring in Bangalore