STYLASTERIDAE 



which the Stolons emerge. Examination of serial sections seems to suggest, that the endoderm in the 

 basal part itself is strongly vacuolated; but this may also be due to the imperfect state of preserva- 

 tion. The protoplasm is here strongly granulated and all indicates, that the albumen cells (cf. Schneider 

 1902 p. 579) form the principal mass of the endoderm in the basal part of the polyp, whilst the 



nutriment cells are in majority in the endoderm of the free wall of the 



zooid. Apart from this localisation of the cell types the gasterozooid agrees 



/ in the whole of its structure with the Hydroid polyp, as is clearly shown by 



the figures (Text-fig. B, PL III, figs. 28 and 29). Whilst the large cnidocysts 



occur in fair numbers in the stolons, the small ones are concentrated in the 



walls of the zooids. In the gasterozooids (PI. Ill, fig. 29) the cnidocysts are 



Text-fig. B. The contracted accumulated densely in the ectoderjn near to the mouth of the polyp and 



gasterozooid of Piiobothrus gy^irely disappear further down on the polyp wall; in the dactylozooids they 



symmelrkus. ec = ectoderm, 



f«= endoderm, m=mouth, are more uniformly distributed in the ectoderm in the whole length ot the 



i = stolones, ^e^=theepi- ^^^^^ ^^^^ become however less numerous near its basis. — To judge from 



the! of the gasteropore, ^p ' 



= mouth of the gastero- Moseley's drawing (1881 PI. VIII, fig. 2) the gasterozooid should have a wide, 



^ ■ cruciform mouth in Pliobflthriis symmetricits. The mouth of the polyp is 



closed in all the specimens examined from the "Ingolf" Expedition and is neither larger nor shaped 



differently from that of the Hydroids generally. The cruciform appearance noted by Moseley is 



clearly due to chance, for the protruding parts of the endoderm in the numerous sections examined 



vary greatly in form, size and number. 



In their di.scn.ssion of the family characters of the Stylastcridae Hick son and England state 

 (1905 p. 13): "The solid scalariform endoderm of the dactylozooids is another very distinctive feature 

 of the group". This does not hold good for Piiobothrus symmetricns] the dactylozooids have a distinct 

 lumen in their centre, as shown in transverse sections (PI. IV, fig. 34). The authors cited maintain, 

 that "the scalariform tissue is clearly a much more differentiated tissue", — naturally, under the tacit 

 assumption, that the dactylozooid is a reduced polyp. This would suggest, therefore, that Moseley's 

 view of Piiobothrus as a primitive genus of the Stylasteridac is correct. 



There is one condition, however, which seems to point in the opposite direction; this is the 

 male gonophores and their structure. Moseley indicated already, that their appearance differs from 

 that found elsewhere in the Stylastcridae. But his drawing (1881 PI. VIII, fig. 3) is in so far misleading, 

 as it gives the impression, that the difference only consists in a denser accumulation of very small, 

 single gonophores. A series of sections through the male gonophores (PI. IV, figs. 40—42) shows ex- 

 tremely aberrant features. Each ampulla contains a single, fairly large, globular gonophore, which is 

 composed of a number of follicle-like portions (pseudofoUicles); the reproductive cells are in extremely 

 different stages of development in the various pseudofoUicles of the gonophore. The spadix as a rule 

 is slightly branched between the pseudofoUicles and sometimes we find, that .slender connecting bridges 

 lead from the stolons of the ampullar wall over to other parts of the gonophore wall than those where 

 its .spadix connects with the stolons of the colony. — Hickson (1891) has demonstrated the forma- 

 tion of a "seminal duct" in Stylasteridac where the male gonophores are of primitive structure. Such 

 a formation can certainly not be found in Piiobothrus symtnctricus. The apex of the compound gono- 



