[ref. ID; 5727 (Foissner & Foissner, 1993)]

Phylum Hemimastigophora Foissner, Blatterer & Foissner 1988
(Class Hemimastigea Foissner et al., 1988 and order Hemimastigida Foissner et al., 1988 have the same characteristics)
Cilia-like flagella arranged in slightly spiraled kineties shorter or as long as body. Single basal bodies with complex system of microtubular roots and posterior membranous sac. Cortex consisting of plicate plates with diagonal (rotational) symmetry; supported by distinct, finely granular layer (epiplasm) in flagellated region and by microtubules either evenly spaced or in discrete groups. Mitochondrial cristae tubular to saccular. Complex extrusomes. Fission in free-swimming condition, symmetrogenic. Incertae sedis within kingdom Protista Haeckel.
Remarks; No left and right side can be distinguished in the Hemimastigophora because of the special diagonal (rotational) symmetry of the cortical plates. The cortex is supported by microtubules and/or by a finely granular layer of moderate electron density. This layer was called epiplasm because of its similarity with the epiplasm of ciliates and euglenids. In some species (H. amphikineta, S. terricola) the outer zone of the epiplasm appears as a heavy osmiophilic sheet. This sheet consists of amorphous material, i.e. lacks any tripartite structure indicative of a membrane. The mitochondrial cristae are irregularly tubular to saccular. One or two tubules in each mitochondrion may be inflated to roundish caverns. Such caverns are present in all species investigated and are therefore unlikely to be fixation artifacts. Disk-like (euglenoid) cristae have never been seen. The Hemimastigophora lack paraxial or paraflagellar structures, flagellar hairs (mastigonemes) and plastids.

Family Spironemidae Doflein, 1916
Small to medium sized (about 10-60 um), vermiform to ovoid, more or less distinctly flattened hemimastigids with more or less distinct anterior constriction producing head-like "capitulum". Two spiraled kineties in more (Hemimastix) or less (Spironema, Stereonema) distinct furrows located at the sites where the cortical plates abut. Basal bodies conspicuously short (about as long as wide), each associated with membranous sac, short microtubular ribbon and long microtubular ribbon extending to or slightly beyond posterior kinetid, where it parallels and sometimes touches short ribbon; microtubular ribbons of same basal body connected by granular material; flagellar basket consisting of 9 filamentous (non-microtubular) arms (transitional fibres) originating from proximal end of basal body; transitional plate thick and concave, transitional cylinder incospicuous. Single nucleus with prominent central nucleolus persisting throughout division. Contractile vacuole near posterior end body. Extrusomes bottle-like, consisting of cylindroid posterior compartment and rod-like anterior compartment. Food uptake at anterior end; defined oral structures, however, not recognizable. Three genera (Spironema, Hemimastix, Stereonema) with a total of five species.

Type genus Spironema Klebs, 1893
Remarks; Whether hemimastigophoran protists with more than two kineties exist is not known. If so, they should be separated at family level. The flagella arise from single pits in Spironema and Stereonema (indistinct furrow) or from a continuous groove in Hemimastix (distinct furrow). The complex extrusomes are remarkably similar in all genera though differing in details. This and the very similar fine structure of the kinetids and the mitochondria indicate a close relationship confirming the classification of Hemimastix amphikineta as a member of the Spironemidae. Although defined oral structures are not recognizable, the Hemimastigophora feed on bacteria and/or flagellates. Food uptake has been observed at the anterior end, but the exact location of ingestion is not known. Contrary to our earlier idea, we now regard food uptake in the centre of the capitulum as rather unlikely because of the continuous epiplasm. Ingestion in the anterior regions of the flagellar furrows, at sites which lack the epiplasm, is more likely and agrees with the obervations by Klebs (1893). Dictyosomes are always in the anterior cell portion and may provide food vacuole membranes.

Phylogenetic relationships of the Hemimastigophora
Foissner et al. (1988) discussed extensively possible relationships of the Hemimastigophora with other protistan phyla. Thus, we focus our discussion on a few points mainly emerging from the new data presented in this paper. Doubtless, the Spironemidae possess a unique combination of characters not found in other protistan phyla. Some affinities eixst to the Ciliophora (microtubular associates of basal bodies = infraciliature) and to Stephanopogon (kineties composed of single, short basal bodies connected, however, by microfibrillar traces; futhermore, mitochondrial cristae are dist-like) (Lipscomb and Corliss 1982, Patterson and Brugerolle 1988). Foissner et al. (1988) suggested a relationship with the Euglenida mainly because of the diagonal (rotational) symmetry of the cortex (compare Mignot 1966, Pochmann 1953). In fact, it is easy to model a Hemimastix by combining two euglenid cortical strips. The idea of a relationship between euglenid and hemimastigophoran flagellates is strongly supported by our obsevations on Spironema terricola. The movement of this organism is impressive and very similar to that found, e.g. in Euglena viridis and Peranema trichophorum. Euglenoid movement was already emphasized by Klebs in Spironema multiciliatum (Klebs 1893) and the resulting twisting of the body was mentioned and adequately illustrated in Spironema goodeyi. The euglenid movement ("metaboly") is due to the sliding of cortical strips relative to each other (Suzaki and Williamson 1986). In euglenids, metaboly is most pronounced in genera with many strips and less distinct of absent in species with few pellicular strips (Leedale 1967, Triemer and Farmer 1991). Although the precise mechanism of metaboly in the Hemimastigophora remains to be elucidated, our observations strongly suggest that the number of strips per se is not a causative agent for euglenid movement and should therefore not be used to separate euglenids from other taxa, like Hemimastix, as recently suggested (Triemer and Farmer 1991). Metaboly, but not the typical euglenid movement, is reported to occur in some kinetoplastids (Vickerman, pers. commun. and Vickerman 1989), which are regarded as relatives or ancestors of euglenids by many specialists (Brugerolle 1985, Corliss 1982, Kivic and Walne 1984, Triemer and Farmer 1991, Vickerman 1989, Walne and Kivic 1989). Their cell membrane (especially that of the trypanosomes) and that of Stephanpogon is supported by a monolayer of microtubules (Brugerolle et al. 1979, Lipscomb and Corliss 1982, Patterson and Brugerolle 1988, Vickermann 1989). This is similar to the microtubule layer beneath the epiplasm-free regions of the Hemimastigophora, especially of Stereonema and Spironema. A noteworthly difference is, however, that the cortical microtubules of the Hemimastigophora do not orginate near flagellar roots or basal bodies as in euglenids kinetoplastids (Brugerolle et al. 1979, Willey and Wibel 1987), but from organelle-free regions of the cytoplasm (tail) or near the cortex. Some bodonids and some euglenids possess extremely long extrusomes similar to the anterior compartment of the extrusomes of the Spironemidae (Brugerolle 1985). Taken together, the Hemimastigophora share more characters with the euglenid-kinetoplastid assemblage than with any other protistan taxon. The most important synapomorphy is the diagonal symmetry of the cortical plates. It is highly unlikely that such a complex character, resulting in a special kind of movement ("euglenoid metaboly"), evolved twice. If we accept Hennig's (Hennig 1982) phylogenetic concept, this complex synapomorphy is sufficient to establish a sister group relationship between euglenids and hemimastigophorans. The ancestral euglenid is proposed to be a phagotrophic flagellate with two basal bodies and a rigid cortex consisting of a few strips (Triemer and Farmer 1991). Apparently, the Hemimastigophora fit most of these characters. Recently, Cavalier-Smith (1991) separated Hemimastix widely from the Euglenozoa because of its tubular mitochondrial cirstae. In our opinion, this character is much weaker (see, for instance, the kinetoplastids and Jacoba, which have both tubular and disc-like cristae (Patterson 1990, Vickerman 1989) and O'Kelly [The genus Jakoba, a remarkable flagellate: structure, reproduction, and possible phylogenetic significance. J. Euk. Microbiol., 40:7A.]) than the unique cortical architecture relating euglenid and hemimastigophoran flagellates. For the moment, however, we keep separate the Hemimastigophora from the Euglenozoa mainly because of the kineties with their distinct "infraciliature", the different shape of the mitochondrial cristae, the lack of flagellar hairs and paraxial rods and the absence of a well-marked flagellar pocket. These differences indicate a long-lasting, separate evolution of the Hemimastigophora. Another classification was suggested by Karpov (1990). He assigns Hemimastix to the order Apusomonadida Karpov & Mylnikov, mainly because he interprets the osmiophilic sheet in the epiplasm of Hemimastix as an additional membrane like that found in Apusomonas. This is, however, a misinterpretation, since this sheet has no tripartite structure. Furthermore, it is absent in Stereonema geiseri, indicating that it is indeed a component of the epiplasm in Hemimastix and Spironema. In our opinion, Apusomonas is very different from the hemimastigophoran flagellates, since it has no cortical plates (and thus no diagonal symmetry), only two highly specialized flagella, and a very different system of flagellar roots.