Organism of the Siplionojjhora. 197 



the distinction of two subfamilies as Cannophysidai and 

 Linopliysidse upon differences which perhaps justify generic 

 separation. The same thing appHes to the splitting of the 

 genera of Agalmidge so far as in their foundation the form of 

 the tentilla is exclusively taken into account [Agalmopsis — 

 Lychnagahna ; Halistemma — Cujndifa ; Anthemodes — Caneo- 

 laria ; Agalma — PhyUojjhysa^ Stephanoinia — Crystallodes). 



Further, it seems to me quite unjustifiable to establish a 

 special order of Siphonophora for the remarkable deep-sea 

 genera Stephalia {Stephonalia) , Auralta, and Rhodalia, as 

 these forms possess the pneumatophore of the Physophorida^ 

 (Physonectaj) and have only acquired the character peculiar 

 to them and by which they take tlieir place as a special group 

 of Physophoridse by the union of the proximal section of the 

 pneumatophore with an air-discharging apparatus (auro- 

 phore). That the peculiar apparatus designated an aurophore 

 has been produced by the transformation of a nectocalyx is 

 not only not proved, but is even very improbable, as we can- 

 not very well see how a nectocalyx could have got upon the 

 dorsal line of the stem, which is always destitute of buds. 

 Even if this remarkable pneumoduct should be superinduced, 

 in analogy with the foundation of the nectocalyx, by a bud- 

 like elevation of the two cell-layers of the stem with subse- 

 quent growth of the entoderm and invagination of the sur- 

 rounding entoderm, this would by no means prove that it was 

 actually produced by transformation of a nectocalyx, but it 

 would be much more justly interpreted as a special differen- 

 tiation of the wall of the stem at the air-funnel of the pneuma- 

 tophore in connexion with the necessity of the escape of air. 

 However, even in the first case there would be no reason for 

 the establishment of a special order. 



Another much heavier criticism relates to the classification 

 of the Calycophoridse (Calyconectge), under which the 

 Eudoxidge and Ersgeidte with their genera and species figure 

 as distinct families side by side with the Monophyida; and 

 Diphyidge. It is, in fact, a fundamental offence against the 

 idea of a natural system constructed upon a phylogenetic 

 foundation to separate the sexual generations which have 

 become independent from the generations which produce 

 them and to treat them as distinct species of distinct genera 

 and families, to be arranged and enumerated as equivalent to 

 the corresponding categories of the nursing generations. No 

 fewer than 25 species, 8 genera, and 2 families in consequence 

 occur twice over and under two denominations. In point of 

 fact such a duplication of equivalent categories would con- 



