20 STVLASTHRIDAE 



corallines, where they occur, are capitate or rather knobbed at their tips . Closer comparison of his 

 drawings and of the figures later authors have given indicates, however, that the Stylasteridac do not 

 have capitate tentacles like the Corynidae, even though the cnidocysts are accumulated more densely 

 in the distal part of the tentacles. It has been pointed out several times in the foregoing, that the 

 tentacles in the northern Stylasteridae fully agree in their structure with the thread-shaped tentacles 

 of the Bougaiwvilliidae. On the other hand, the Milleporidae have typical, capitate Corynid tentacles. 

 This suggests that the Hydrocorallines have a diphyletic origin and that the excretion of a calcareous 

 skeleton in the two groups is purely a character of convergence. This witnesses further to the 

 correctness of the separation of the Hydrocorallines into two families, as is done by Hicks on and 

 England (1905 p. 1.). On the other hand, we are not entitled to regard the two groups as sub-orders 

 we should rather consider them as two highly specialised Hydroid families. Hickson and England 

 take the Milleporidae as Hydroids, whilst the}- regard the Stylasteridac as most nearly related to the 

 Trac/wmcdusae; what support they may have for this last view, does not appear anywhere in their 

 works; the consequence is, that Hickson in his contribution to The Cambrigde Natural History 

 (1906) discusses the two families at widely separated places. — The structure of the gasterozooid in 

 the Stylasterid genera as in Sporadopora and still more Errina agrees fully with the polyp of the 

 Bougainvilliidae\ excluding the gasterostyle, which phylogenetically must be of fairly recent origin and 

 which does not occur either in all Stylasterids, the gasterozooids of the Stylasteridae fully agree with 

 the polyps of Clathrozoon. This indicates, that the latter genus contains the nearest allies of the 

 Stylasteridae among the Hydroids. 



The second main character, the polymorphic development of the polyps in the Stylasteridae, 

 we find already indicated in such Bougainvilliidae as Hydractinia; in several species of these we find 

 tentacle-less »feelers« and tentacle-bearing nutritive individuals. In the Hydractinia species this 

 distinction is not considered a useful mark of generic separation, nor can it be considered so important 

 a feature in the Stylasteridae, that it can form the basis for a separation of the group into a special 

 order, even if the dimorphic development of the somatic individuals has become a constant character. 



We have hitherto directed attention exclusively to the somatic individuals, assuming that they 

 give the most important supports in judging of the phylogenetic conditions of the Coeleuterates. — 

 Hickson (1891) through his evidence of a medusa generation in Millepora has produced the last 

 incontrovertible proof of the Hydroid nature of this genus; the medusa seems to be an undoubted 

 Anthomedusa and thus shows clearly the close relationship with the Corynidae. On the other hand 

 the Stylasteridae have sessile gouophores. Moseley (1881 p. 93) maintains that the gonophores 

 of the Stylasteridac are .adelocodonic* ; but he regards the spadix of the female gonophores as an 

 organ (the trophodisc ) special to the Stylasteridae. Hickson later (1888 and 1891) has examined 

 the gonophores more closely and supports Moseley's view, that the trophodisc is a special formation 

 in the Stylasteridae which is no direct parallel to the spadix of the hydroid gonophores. 



If we consider the manifold nature of the development and organisation shown by the gono- 

 phores of the Hydroids, the special character of the trophodisc and gonophores of the Stylasteridac 

 becomes extremely doubtful. In his excellent studies on the hydroid gonophores (1910) Kiihn has 

 shown, that the medusa reduction may be so complete, that even the endocodou may disappear; thus 



