The flagellum has been shown by Fischer to have one of two forms: either it is whip-like, the stick, alone visible in the fresh specimen, being seen when stained to be continued into a long lash, hitherto invisible; or the whole length is fringed with fine ciliiform lateral outgrowths. If single it is almost always protruded as a tugging organ ("tractellum");[120] the chief exceptions are the Craspedomonads, where it is posterior and acts as a scull ("pulsellum"), and some Dinoflagellates, where it is reversible in action or posterior. In addition to the anterior flagellum there may be one or more posterior ones, which trail behind as sense organs, or may anchor the cell by their tips. _Dallingeria_ has two of these, and _Bodo saltans_ a single anterior anchoring lash, by which they spring up and down against the organic débris among which they live, and disintegrate it. The numerous similar long flagella of the Trichonymphidae afford a transition in the genus _Pyrsonympha_ to the short abundant cilia of _Opalina_, usually referred to the Ciliate Infusoria. {115}An undulating membrane occurs, sometimes passing into the flagellum in certain genera, all parasitic, such as _Trypanosoma_ (incl. _Herpetomonas_), _Trichomonas_, _Hexamitus_, and _Dinenympha_. In some cases the flagellum (or flagella) is inserted into a definite pit, which in allied forms is the mouth-opening. The contractile vacuole is present in the fresh-water forms, but not in all the marine ones, nor in the endoparasites. It may be single or surrounded by a ring of minute "formative" vacuoles or discharge into a permanently visible "reservoir." This again may discharge directly to the surface or through the pit or canal in which the flagellum takes origin (_Euglena_). The "chromatophore" may be a single or double plate, or multiple.[121] In the peculiar form _Paramoeba_ the chromatophore may degenerate and be reproduced anew. It often encloses rounded or polygonal granules of uncoloured plasma, very refractive, known as "pyrenoids." These, like the chromatophores, multiply by direct fission. The "reserves" may be (1) fat-globules; (2) granules of a possibly proteid substance termed "leucosin"; (3) a carbohydrate termed "paramylum," differing slightly from starch (see p. 95); (4) true starch, which is usually deposited in minute granules to form an investment for the pyrenoid when such is present. A strongly staining granule is usually present in the plasma near the base of the flagellum. This we may term a "blepharoplast" or a "centrosome" in the wider sense.