STYI.ASTERIDAE 



phore is prolonged (PI. IV, fig. 42) but seems never to reach the surface of the colony. To 

 judge from serial sections it opens out into one of the surroxinding stolon canals of the calcareous 

 skeleton at the side of and not in the stolon itself. The available material however is not sufficiently 

 well-preserved to enable these conditions to be closely studied, nor does it give any clear picture of 

 the structure or development of the opening of the gonophore. 



It would be of great interest to clear up the origin and development of these compound gono- 

 phores. Probablv we have here a secondary coalescence of the whole gonophore complex which we 

 find collected into the single ampulla in most of the other Stylasferidae. The branched spadix and 

 the variating stage of development of the reproductive cells in the single pseudofoUicles point in this 

 direction. Investigation of the Sfylaster species shows, that the male gonophores in an ampulla are in very 

 different stages of development and thus function over a long period; the successive maturation, which 

 obvioush- must be of great importance for the species, is thus retained in Pliobofhnis symmctricus after 

 the gonophores have become fused into a single complex. Unfortunately the material has not per- 

 mitted any close study of the development of the gonophore in the present species. — I was unable 

 to find other developmental stages of the female gonophores than those already known from the 

 work of Moseley. Thus nothing was found which differed from his description. — 



PUobothrus sytmnetriciis was first described by Pour tales (1871) from the Florida reefs, 

 where it seems to be common between 180 and 300 m. in depth. The species has only once been 

 met with in the Northern Atlantic, where P. M. Duncan (1874) mentions it from the cold area of 

 the Faeroe Channel. 



Sty 1 aster Gray.^ 



The usually fan-shaped colonies have the zooids collected into closed cyclosystems, which show 

 no trace of opercula. Both the gasteropores and the dactylopores are provided with styles; the dactylo- 

 styles however may be rudimentary. 



On the basis of this diagnosis we must also include the genus Allopora Ehrenberg in Stylastrr. 

 Hickson and England (1905 p. 6) rightly point out how small and doubtful the characters are, 

 which are said to distinguish the two genera. In reality there are quite even transitions between the 

 genera. None have shown this more clearly than Hickson and England, wlio following the 

 proposal of Studer group the species into four divisions: "A. Cyclosystems on lateral sides of branches 

 onl)-; B. Cyclosystems on lateral sides of branches and a few on the surfaces; C. Cyclosystems evenly 

 distributed over the surfaces of the branches, and D. Cyclosystems on the anterior surface of the 

 branches only." The first two groups stand extremely near to each other and should be embraced 

 within one main group or subgenus Eustylastcr\ the last group has just the arrangement of the 

 cyclosystems which is characteristic of Allopora; species like Sfylaster diverge 11 s Marenzeller and Allo- 

 pora rosacea Greeff obviously occupy intermediate positions between our northern Allopora species with 

 its hardly prominent cyclosystems and those Stylaster species which have very prominent, almost 

 stalked cyclosystems. For group D. the old generic name Allopora should be retained. 



The genus Stylaster is represented in the Northern Atlantic by three species, which have been 

 confounded several times. When the ampullae are strongly developed the colonies in our Fm- 



