REPORT OF THE MALARIA EXPEDITIONT. 9 
The morphology of these Ha;mamoebidae presented 110 differences from the accepted 
standards. One of us has been inclined to admit two varieties of H. vivax, one with fine lisiht 
brown pigment, and one with fine black pigment. Both were observed in Freetown ; but the 
latter only in one case. 
We were much struck by the absence or paucity of crescents (gametocytes) in cases of 
H. prcecox. This fact will be discussed later [^paragraph 7,9]. 
We found H. dan'tlctuskii of the normal type in a few small birds ; but H. rclicta was not 
seen by us. 
We wished to examine the blood of a large number of the civil population in order to 
estimate the percentage of them in which the parasites are to be detected ; but found sucli an 
investigation impracticable. 
8. Methods for ascertaining Definitive Hosts. — The long researches of one of us 
in India, followed by those of Koch, Daniels, and the Italian investigators, have given us a very 
exact knowledge of the life-history of the Haemamoebidae in gnats, and have shown us how to detect 
them in the insects with ease and certainty. It has been noted that in inhospitable species of 
gnats the ingested parasites perish within the stomach cavity ; whereas in hospitable species the 
zygotes escape from that cavity and develop in the tissues, ultimately giving rise to blasts which 
are found in the juices and salivary glands of the insect. Hence to incriminate a given species of 
insect with regard to the human Hasmamoebidfe, it is necessary only to find the parasites within 
that species. In short, the method to be adopted is precisely that used in ascertaining alternative 
hosts for any other metoxenous parasite, such as Cestodes and Filaridae, for instance. Care only 
is required to determine for certain that the parasites found in the gnats were derived from human 
beings, and not from other vertebrates. It is not necessary actually to infect human beings from 
infected gnats, because the general law that infection can be so produced has already been 
sufficiently established. 
Hence a given species of insect is incriminated simply if we succeed in finding the 
parasites — 
(1) in insects bred from the larva and fed on infected persons. 
(2) in any insects fed on infected persons — provided that the zygotes found in the insects 
are of the proper stage of growth ; or 
(3) in a large percentage of insects caught in a chamber where an infected person lives — 
provided that the circumstances exclude the possibility of the insects having become 
infected from other vertebrates. 
9. Definitive Hosts in Freetown. Our first object being the searcli for definitive liosts, 
we propose to record our operations in this connection in some detail. 
For the reasons given in paragraph 22, suspicion was already strongly excited against all 
species of gnats of the genus Anopheles ; and, accordingly, we began by searching for these insects. 
Two species were found almost at once, under the following circumstances. 
B 
