that is, until it is outside the body I concluded that the function, then unknown, of the flagellum lay outside the human body, and that the flagellated body was the first phase of the extracorporeal life of the malaria parasite. As the parasite whilst in the circulation is always enclosed in a blood-corpuscle, and is therefore incapable of leaving the body by its own efforts, and as it is never, so far as known, extruded in the excreta, I concluded that it is removed from the circulation by some blood-sucking animal, most probably by some suctorial insect common in the haunts of malaria. This blood-sucker I believed to be the mosquito, an insect whose habits seemed adapted for such a purpose, and whose distribution conformed to the well-ascertained habits of malaria. Further, basing my argument on what I had shown to be the fact in the case of Filaria bancrofti, and on the peculiarities of the distribution of malaria, I reasoned that only particular species of mosquito were capable of subserving particular species of malaria parasite.
Ross's observations.—Ross, to whom I suggested the investigation of this hypothesis, first (1895) demonstrated the fact that when crescent-containing blood is ingested by the mosquito a large proportion of the crescents rapidly proceed to microgamete formation and to the emission of microgametes . In l897 he showed that, in a particular species of mosquito fed on malarial blood, living and growing malaria parasites containing haemozoin are to be found embedded in the stomach wall of the insects. Early in 1898 (as announced by me at the meeting of the British Medical Association, in Edinburgh, in July, 1898) he showed that if a particular species of mosquito be fed on the blood of plasmodium-infected birds, the parasite, which both in habit and structure closely resembles the malaria parasite of man, enters the stomach wall of the insect, grows and sporulates there, and that the resulting sporozoites subsequently enter the salivary gland of the insect, and that the insect is then capable of infecting by its bite other birds. Ross further showed that only particular species of mosquito could subserve the avian plasmodium in this way, and that the particular mosquito
