In the obscurity of Amoy, South China, Patrick Manson first recognized that a bloodsucking arthropod can harbor--and presumably transmit--organisms of human disease: in particular, that the house mosquito is an intermediary of the filarial parasite Wuchereria bancrofti. Manson published his find in 1878 and within two years discovered that the parasite's "embryos" (microfilariae) exhibit "nocturnal periodicity," i.e., they absent themselves from the blood during the day and reappear at night coincident with the vector's greatest biting activity. The historical sequence of Manson's two findings have been confused and reversed in the literature, and in this paper evidence is adduced that the sequence given above is correct. Until the turn of this century, zoologists thought that mosquitoes took one blood-meal and died on water a few days later. Manson theorized that filarial larvae escaped from the mosquitoes into water and that people ostensibly infected themselves by drinking the contaminated water. When, however, infective larvae were found in 1900 in the mouthparts of mosquitoes, the concept that transmission occurred by bite (malaria transmission was by then understood) proved unavoidable. Nevertheless, about 70 years after the demise of Manson's old concept, workers succeeded in transmitting several kinds of W. bancrofti-related filarial organisms by mouth to gerbils. This finding served to support Manson's speculations of 1878-1900, altered the views of the parasite group to which the filariae belong, and now raises the difficult question of whether oral infection of humans does or does not occur naturally.