CEETAIN STICHODACTTLIK^ TO THE MADKEPORAUIA. 643 



of the oral aperture. As shown in microscopic sections, they are 

 not to be confuted with the position of attachment of the 

 mesenteries. 



The presence or absence of oesophageal or gonidial grooves in 

 Actiniariaa polyps cannot now be regarded as a character of 

 much pbylogenic importance, seeing that so many exceptions and 

 not well-defined cases are recorded in widely different groups. 

 Their association with the directives is also incousiant. In 

 Actinoporus elegaiis, Duch,, I have found two j)airs of directives 

 and only one oesophageal groove. This is all the more noteworthy, 

 as in this instance the single groove is exceptionally large and 

 clearly defined by considerable histological modification. 



Corynactis, Ricordea, and Rhodactis are devoid of gonidial 

 grooves throughout the length of the stomodseum, and it is 

 significant that apparently no grooves are known to occur in any 

 recent coral polyps, though there is little doubt that the fossulse 

 in Palgeozoic corals are the correlatives of the presence of gonidial 

 grooves and shorter directive mesenteries in these ancient types 

 of coral polyps. 



"With regard to the stomodaeum in Fimgia, Bourne (1887, p. 17) 

 observes: — "I was unable to trace the existence of gonidial 

 grooves (siphonoglyphes) at its ends, though they no doubt exist. 

 When alive the animal constantly closes the middle portion of 

 its mouth, leaving small apertures at the extreme ends through 

 which currents of water pass in and out." This action is well 

 known in anemones, and I have also observed it in living corals ; 

 it does not necessarily indicate that true gonidial grooves are 

 present. 



Mesenterial Pilaments. 



Commenting upon Hertwig's suggestion in the ' Challenger ' 

 Report that at some time it may be found not possible to separate 

 Zoantharian families without from those with a skeleton, 

 H. V. Wilson (1890, p. 3) remarks : — " A strong argument 

 against this view of the Madreporaria is, I think, afforded by a 

 study of the mesenterial filaments. Porous and aporous corals 

 alike have simple filaments, but actinise have trifid filaments." 

 Since this was written several Actinige have been described in 

 which the mesenterial filaments are simple throughout. The 



