(1) Herpetomonas jaculum is a parasite of Nepa cinerea, occurring in the alimentary tract of its host.

(2) The life-history of the parasite may be conveniently divided into three stages, the pre-flagellate, flagellate and post-flagellate stages, which gradually merge into one another.

(3) The movements of the parasite are less flexible than those of Crithidia, as Herpetomonads have no undulating membrane. The flagellum is the most active agent in effecting motion.

(4) The pre-flagellate stages of the parasite (Pl V, Figs. 1–18) are best observed in the crops of nymphs of Nepa cinerea. The parasites at first are oval (Figs. 1–6). They vary in size from about 4μ to 5μ long and from 2μ to 2·5μ broad. They show nucleus and blepharoplast, and may divide longitudinally before flagella are acquired (Figs. 2–4). The flagellum of each parasite arises from a region near to the blepharoplast but not directly from it (Figs. 5–9).

(5) The flagellate stage of the organism (Pl. V, Figs. 19–36) is that best known. H. jaculum is from 13μ to 33μ long and from 1μ to 4μ broad, the size varying according to the recency or otherwise of longitudinal division. Myonemes (Figs. 20, 21, 38) are present on the body. The flagellum is at least as long again as the body. The nucleus contains a number of grains of chromatin (Figs. 28, 34), sometimes in the form of eight large grains (Figs. 34, 36), sometimes as very fine granules (Figs. 22, 31). The blepharoplast is in the anterior, pre-nuclear, region of the parasite, and is usually rod-like (Figs. 19, 20, 21). The single flagellum (Fig. 19) arises near it but not from it. A basal granule (Figs. 22, 33, 34) is present at or near the origin of the flagellum. Chromidia are present as scattered granules in the body (Figs. 28, 32, 33).

(6) The post-flagellate stage is the form assumed by the parasite for life outside the body of the host. Preceding encystment, the organism divides twice longitudinally, giving rise to four daughter forms (Pl. V, Figs. 51–54) each of which ultimately loses its flagellum, rounds itself off and forms a cyst (Figs. 57–68). These cysts are from 2·5μ to 4·5μ long and from l·4μ to 2·6μ, broad. They occur in the rectum of Nepa cinerea and are voided with the faeces, being ingested later by other bugs.

(7) Longitudinal division is the common method of multiplication of H. jaculum. The flagellum may divide precociously, but usually division is initiated by constriction of the blepharoplast (Pl. V, Fig. 37) almost simultaneously with division of the flagellum and followed by that of the nucleus (Figs. 39, 40). A split occurs (Figs. 40–42) and the active movements of the two flagella aid in the divergence of the daughter organisms (Figs. 43, 44), which ultimately separate.

(8) I have no evidence whatever for ascribing sex to any form of Herpetomonas, but consider the occurrence of long and short and of thin and stout forms to be explicable as the results of growth and division. Also, I have shown experimentally that richly granular protoplasm is the result of a physiological condition and is not necessarily fixed as an attribute of the female sex.

(9) One mode of infection has been proved experimentally in the laboratory and also observed at the breeding grounds of the Nepa. Cysts voided in infected faeces are swallowed by other Nepa in the adult and nymphal stages. The crops of such nymphs on dissection were found to contain cysts, whereas no flagellates were present in other parts of the gut, so that primary infection occurred here in the crops of the nymphs.

Cannibalism of Nepa cinerea whereby other Nepa are devoured is also responsible for the spread of H. jaculum.

I have no definite evidence of hereditary infection, although I have found flagellate and post-flagellate forms (Pl. V, Figs. 46–49) in the ovaries of the host. The parasites were not found in the eggs. The occurrence of parasites in the ovaries may be regarded as a stage in the evolution of hereditary infection.

(10) Various environmental effects have been studied, the most important observation being that fresh food appears to stimulate the parasites and to cause their rapid division.

(11) The generic name Herpetomonas should be retained, as originally constituted, for parasites having but one flagellum and no undulating membrane (see pp. 383 et seq.).

(12) I wish to record the occurrence of a new species of Herpetomonas, H. vespae, from the alimentary tract of the hornet, Vespa crabro.