Bn, basal nematocyst knob; Cc, circular canal; Db, dorsal branch of tentacle; Es, eye spot; G, gaster, Gn, gonad pouch; Gf, gastric pouch; M, manubrium; N, nematocyst knob; Rc, radial canal; T, tentacle; TN, terminal nematocyst knob; V, velum; VB, ventral branch of tentacle.
species recorded by Gilchrist, i.e. C. vallentini, charcoti, hodgsoni, kerguelense and capensis and adds the Australian form as C. haswelli n.sp. Lengerich (1923) follows Vanhöffen (1911) and considers all the known species of Cnidonema as referable to C. vallentini.
Edmondson (1930) records four new species of Eleutheria (Northern Hemisphere genus) from Hawaii and states that “the species of the Southern Hemisphere, if the several forms are to be considered identical, is more closly allied to the Hawaiian ones than are the Hawaiian to those of Europe. An affinity between Cnidonema vallentini and the Hawaiian species is seen in that all possess clusters of nematocysts on the dorsal branch of each mature tentacle in addition to the capitate group at its extremity.” The chief characters on which Edmondson erects his species are the position of the accessory stinging batteries and the form of the tentacle. Weill (1937) considers this insufficient evidence to separate the Northern Pacific species, i.e. Edmondson's species of Eleutheria, from the Australo-antarctic (species of Cnidonema). Weill further considers the form he found at Bermuda definitely Cnidonema without corresponding exactly with any of the numerous species of the genus. He thinks it premature to call the Bermudan specimens a new species as those already existing are in his opinion justifiably questioned. Weill refers the Bermudan specimens, which he considers juvenile, to the type species C. vallentini. The external form of the Cnidonema figured by Weill is very similar to the juvenile form of the present material, except that in the New Zealand specimens asexual budding has not been observed.
From the evidence of my material I agree with Vanhöffen (1911), Lengerich (1923), and Weill (1937) in uniting the various species of Cnidonema under the type species C. vallentini. It should be noted, however, that C. vallentini is the only southern species that has been described as having the gonad entirely above the stomach. This point cannot be checked until further material is described from Stanley Harbour, Falkland Islands. As all the other species agree in not having the gonad above the stomach and differ mainly in tentacle form and structure, which is a feature that has been shown to change within the life history, it seems best at present to conclude that the southern species are all identical and referable to C. vallentini. Gilchrist (1919 appears to be the only writer up to the present who has observed the hydroid of Cnidonema. The outstanding feature from all descriptions of the genus is its extreme variability. The present paper shows how considerably the medusa changes its form when passing from the juvenile to the adult stage. Great difficulty is therefore created for the worker who may have to depend for his description on a few specimens probably of one phase of the life cycle only. For instance, some forms in the life cycle of the New Zealand specimens (particularly features in relation to the number of accessory stinging batteries and tentacle form) correspond very closely indeed with three of Edmondson's species, namely, E. oahuensis, E. bilateralis, and E. alternata. The probability is that the Hawaiian species are stages in the development of the genus Cnidonema.
C. charcoti and C. hodgsoni are separated as species because C. charcoti has radial canals that tend to anastomose, otherwise the two forms are identical. As branching of the radial canals is a feature of some of the adult forms of the New Zealand specimens, it seems probable that these two species at least are identical one with the other and with the New Zealand specimens. The present writer feels that it is not sufficient to erect species on features that vary within the life cycle, as, for instance, the position and number of nematocyst knobs on the aboral portion of the tentacle (Edmondson's species) or the length of the aboral and oral branch of the tentacle, i.e. the tentacle form (Gilchrist and Briggs). Cnidonema appears to parallel the closely allied genus Cladonema. Weill (1937) gave a full review of the three existing species of Cladonema and came to the conclusion that Cladonema radiatum, the type species, was cosmopolitan and that the existing species were geographical races.
The present material shows a novelty apparently not observed by other workers. In all specimens where satisfactory observations could be made, a ring of five or six stinging knobs can be seen arranged around the manubrium. It may be possible that other workers did not observe these stinging knobs because of the difficulty of recognising them in preserved material and even in fresh material if the manubrium is in rapid motion or at all contracted. Apart from this, the specimens fall well within the wide range of C. vallentini. In ordinary circumstances this novelty might be considered sufficient to establish a new species, but because of the extreme variability of the specific characters and the difficulty of observing the knobs under the circumstances described above, the writer does not feel justified in erecting a new species on this feature alone. If in the future oral stinging batteries should prove to be a novelty in the genus, it may be necessary to establish a new species.