NANOMIA CARA. 213 



with the exception of one root, which is positive, and this would repre- 

 sent the only remaining spheromere, that of the Deckstiick. 



If this view is correct, it is evident that the opinion of Huxley, who 

 considers the scale as homologous to the urn of Campanularians, can- 

 not be sustained. I have already hinted at the similarity of the em- 

 bryonic Nanomia, in the stage of Fig. 347, when it consists of only the 

 large Polyp and the float, with the early stages of the fixed Hydrarium 

 of Melicertum, where we have at first a single Polj^p, from which are 

 developed, by budding, the branches and the other kinds of individuals 

 of the community ; supposing this community, instead of fixing itself, 

 as it does, to remain movable, the base of the stem to expand into a 

 float and become separated from the main cavity, we should have a 

 Siphonophore. The discovery by McCrady and Stimpson of the float- 

 ing Hydrarium of Nemopsis and Acaulis, where the Medusa3 are closely 

 related to genera the Hydrarium of which is always fixed, reduces still 

 further the distinction which has been made of Polypi Nechali. And 

 Avhen we find that there are genuine Medusae (Dysmorphosa) which 

 for four generations reproduce themselves by budding from the jDro- 

 boscis, exactly in the same way in which we find additional individuals 

 arismg along the walls of the original Polyp among Physophoridte, we 

 are at a loss to find any distinctions to separate the Siphonophores 

 from the true Hydroids, and we cannot consider them as anything 

 but floating Hydroid communities. 



Massachusetts Bay, Naliant, and Newport, E. I. (A. Agassiz). 



Cat. No. 365, Nahant, Mass., September, 1862, A. Agassiz. 



Museum Diagram, No. 27, after A. Agassiz. 



