A further divergence of opinion arises from differences in the interpretation of the persons composing the colony. It is possible to regard the cormus (1) as a colony of medusa-persons, (2) as a colony of polyp-persons, (3) as composed partly of one, partly of the other. It is sufficient here to mention briefly the views put forward on this point by C. Chun and R. Haeckel. Chun (HYDROZOA [1]) maintains the older views of Leuckart and Claus, according to which the cormus is to be compared to a floating hydroid colony. It may be regarded as derived from floating polyps similar to _Nemopsis_ or _Pelagohydra_, which by budding produce a colony of polyps and also form medusa-buds. The polyp-individuals form the nutritive siphosome or trophosome. The medusa-buds are either fertile or sterile. If fertile they become free medusae or sessile gonophores. If sterile they remain attached and locomotor in function, forming the nectosome, the pneumatophore and swimming-bells. Haeckel, on the other hand, is in accordance with Balfour in regarding a Siphonophore as a medusome, that is to say, as a colony composed of medusoid persons or organs entirely. Haeckel considers that the Siphonophores have two distinct ancestral lines of evolution: 1. In the _Disconanthae_, i.e. in such forms as _Velella_, _Porpita_, &c., the ancestor was an eight-rayed medusa (_Disconula_) which acquired a pneumatophore as an ectodermal pit on the ex-umbrella, and in which the organs (manubrium, tentacles, &c.) became secondarily multiplied, just as they do in _Gastroblasta_ as the result of incomplete fission. The nearest living allies of the ancestral _Disconula_ are to be sought in the _Pectyllidae_. [Illustration: After Haeckel, from Lankester's _Treatise on Zoology_. FIG. 74.--_Stephalia corona_, a young colony. p, Pneumatophore. n, Nectocalyx. l, Aurophore. lo, Orifice of the aurophore. s, Siphon. t, Tentacle.]