20 AMERICAN HYDHOIDS. 



cylindrical, forming a smooth surface ; while those of the gastric cavity are broader and protrude 

 into the cavity. In the oral end of the proboscis both cells layers are thinner, and here the 

 mouth breaks through both layers and establishes communication with the interior. This 

 occurs before the development of the trumpet-shaped proboscis. 



This same writer treats of the phylogeny of the Thecophora and represents the Sertularidse as 

 having been derived from the Campanularidse,^ basing his conclusions exclusively on the method 

 of budding and growth of the colony. With this conclusion the present writer is unable to 

 agree, as it loms counter to that which, to his mind, is much more weighty evidence tending to 

 show that the Sertularidse can not have been derived from the Campanularidse. A more logical 

 argument, as it seems to the writer, can be based on the characters of the hydranth itself. As 

 a general thing the nutritive persons seem much less easily modified than is the general form 

 of the colony or what might be called the skeletal characters — i. e., the perisarc. The Scandi- 

 navian writer, Broch,^ has separated the forms which come under the family Campanularidae, 

 as here used, from all other Calypteroblastea and placed them under a suborder which he 

 calls "Thecophora Proboscoidea " on account of their trumpet-shaped proboscis. That the 

 Campanularidse differ from all the other families of Calypteroblastea more than these latter do 

 from each other seems cA/ident. The constant and striking character of the hypostome is con- 

 fined to this group and the gynmoblastic family Eudendridse. 



That the trumpet-shaped proboscis is a specialization and therefore a departure from the 

 original type is shown in sequence of events in the growth of the hydranth as described by 

 Kiihn in the work referred to," and sketched on a previous page of the present work. In the 

 course of this development the hydranth assumes first the form of hypostome characteristic 

 of the Sertularidse and afterwards takes on the characteristic trumpet shape of the Campanu- 

 laridse. This seems to be a very strong proof that the former can not be derived from the latter, 

 as claimed by Kiihn. If such a derivation were a fact, we would have a reversal of the ordinary 

 procedure in embryological development that would be unexpected, to say the least. 



Referring again to the similarity of the hydranth of the Campanularidge to that of the 

 Eudendridse it may be worth while to caU attention to the fact that in this latter family we have 

 the nearest approach to the production of the hydrotheca that I have seen in the gynmoblastic 

 hydroids. In Eudendrium vaginatum Alhnan the perisarc of the pedicel is produced upward 

 and expanded into a cup-hke form which covers the hydranth body up to a short distance below 

 the tentacle bases. The proximal part of the female gonophore is protected in the same way 

 by an expansion of the distal end of the perisarc covering the pedicel.* In this species the 

 hydranth with its trumpet-shaped proboscis and its striking pseudo-hydrotheca reminds one 

 very strikingly of a campanularian form. The pseudo-hydrotheca, however, seems to be attached 

 to the hydranth throughout the extent of the former, not free as in the Campanularidse. 



In Garveia annulata Nutting we have a similar instance of a pseudo-hydrotheca in con- 

 nection with a hydranth having a dome-shaped proboscis very similar to that found in the 

 Sertularidse. In discussing this species the describer says : ° 



The structure that I have designated above as a " pseudo-hydrotheca" is of considerable morphological interest, 

 for it may throw light on the origin of the hydrotheca. The extension of the chitinous perisarc of the stem over the 

 body of the hydranth appears to be attached to the latter; A true hydrotheca would be formed if the perisarc around 

 the hydranth body should become thicker and detached. 



In the development of the hydranth as worked out by Kiihn and described on a previous 

 page it is seen that the hydrotheca is at first attached to the hydranth body throughout, and 

 later becomes free. If ontogeny here follows the path of phylogeny it seems evident that 

 such species as Eudendrium vaginatum and Garveia, annulata represent the primitive condi- 

 tion of the hydrotheca. 



• ' Alfred Kiihn, Sprosswachstum und Polypenknospung bei den Thecophoren, 1909, p. 465. 

 ^Die Hydroiden der arktischen Meere, 1909, p. 183. 

 'Sprosswachstum und Polypenknospimg bei den Thecophoren, 1909. 

 * See Nutting, Hydroids of the Harriman Alaska Expedition, 1901, pi. 15, figs. 3-6. 

 'Hydroids of the Harriman Alaska Expedition, 1901, p. 167. 



