The mode of infection of the vertebrate in Nature seems to be contaminative, either by its food or through an already existing abrasion or puncture on the surface of its body. Cases in which the flagellate-infected insects have been allowed to suck the blood of vertebrates have proved negative up to the present. In areas where leishmaniases are endemic, an examination should be made of all insects and other invertebrates likely to come into contact with men or dogs, or rats and mice (see below), in order to ascertain if these invertebrates harbour herpetomonads. Preventive measures should be directed against such invertebrates, especially arthropods. Further, it is likely that certain vertebrates, such as reptiles and amphibia (especially those that are insectivorous), may serve as reservoirs of leishmaniases, or, as they should preferably be termed, herpetomoniases. From such reservoirs the herpetomonads may reach man by the agency of ectoparasites or flies, especially such as are sanguivorous.

That vertebrates in Nature can harbour herpetomonads in their blood has been shown by the work of Dutton and Todd (1903) on the herpetomonads of Gambian mice, while the recently published investigations of Fantham and Porter[1261] (June, 1915) on natural herpetomonads in the blood of mice in England have shown that these rodents may be a natural reservoir of herpetomoniasis. The origin of the infection of mice is to be sought in a flagellate of an ectoparasite of the mouse, very probably Herpetomonas pattoni parasitic in various fleas, which protozoön can adapt itself to life in the blood of mice. Herpetomonads were also found naturally in the blood of birds by Sergent (1907). Recently, Fantham and Porter have successfully infected birds with herpetomonads experimentally.

The significance of insect flagellates in relation to the evolution of disease has recently been set forth by Fantham[1262] (June, 1915). The deductions to be made from the occurrence of a herpetomonad stage in Leishmania, especially in L. tropica, in man himself, and of flagellate stages of the so-called Histoplasma capsulatum in man are fully discussed and correlated. It is pointed out that flagellosis of plants (see p. [104]) may possibly be connected with leishmaniasis. The evolution of Leishmania from flagellates of invertebrates is thus traced and the way again indicated for preventive measures against leishmaniasis, as first set forth by Dodds Price and Rogers.

Franchini and Mantovani (March, 1915) have successfully infected rats and mice by inoculation or by feeding with Herpetomonas muscæ domesticæ obtained from flies and from cultures.

It is of great interest to note that the recent observations of Ed. and Et. Sergent, Lemaire and Senevet[1263] (1914) have demonstrated the presence of a herpetomonad flagellate in cultures of the blood and organs of geckos obtained from areas in Algeria in which Oriental sore, due to L. tropica, is prevalent. Phlebotomus flies, which may harbour a natural herpetomonad, feed on the geckos and on men. Hence animals like geckos may possibly act as reservoirs of leishmaniasis. Lindsay[1264] (1914) writes that the parasite of dermo-mucosal leishmaniasis in Paraguay is believed by native sufferers to be conserved in rattlesnakes, and spread by ticks or flies (Simulium) feeding on the reptiles and transferring the parasite to man.

The Transmission of Spirochæta duttoni (see p. [116]).—It is probable that Ornithodorus savignyi acts as the transmitting agent of S. duttoni in places like Somaliland (Drake-Brockman, 1915).[1265]

Spirochæta bronchialis (see p. [122]).—The morphology and life-history of S. bronchialis have been investigated by Fantham[1266] (July, 1915). From researches conducted in the Anglo-Egyptian Sudan, he found that S. bronchialis is an organism presenting marked polymorphism, a feature that has only been determined by the examination of numerous preparations from the deeper bronchial regions of various patients.

S. bronchialis varies in length from 5 µ to 27 µ, and its breadth is about 0·2 µ to 0·6 µ. These variations are due to the processes of growth and division. Many of the parasites measure either 14 µ to 16 µ long, or 7 µ, to 9 µ, the latter resulting from transverse division of the former. The ends show much variation in form, but approach the acuminate type on the whole. The discrepancies in dimensions given by the very few previous workers on the subject are probably the result of the measurement of a limited number of parasites. All such sizes can be found on some occasion during the progress of the disease, when a larger number of spirochætes is examined.

The movements of S. bronchialis are active, but of relatively short duration, when it is removed from the body. The number of coils of the spirochæte is rather an index of its rapidity of motion than a fixed characteristic of the species.

The motile phase of S. bronchialis is succeeded by one of granule formation, the granules or coccoid bodies serving as a resting stage from which new spirochætes are produced. The formation of coccoid bodies and reproduction of spirochætes from them can be observed in life.