216 PROCEEDINGS OF THE NATIONAL MUSEUM vol. loo 



where Wright and Studer, Versluys, and all subsequent modern ij 

 authors have arranged it. 



The radial orientation of the calcareous material in the ellisellid i 

 axis differs markedly from the situation seen in the families Primnoidae 

 and Chrysogorgiidae, in which the calcareous material permeates 

 imbricating, concentric lamellae and there is no evidence of radial 

 arrangement. The accompanying photomicrographs of axis thin sec- . 

 tions permit comparison of calcification as seen in Ellisellidae (pi. 5) ,j 

 with that of Chrysogorgiidae (pi. 4:,c,d) and Primnoidae^(pl.'4,e,/). 



Family Isididae Laniouroux, 1812 



Structure of the colonial axis: Although the axes of Isididae I 

 differ widely from those of Ellisellidae in their articulation of alter- i 

 nating horny and calcareous segments, they are similar in having the a 

 calcareous material radially oriented. '| 



In Isis hippuris Linnaeus, the calcareous fibers are united into 

 bundles, much like the sclerodermites of stony corals, radiating out- 

 ward from the central core (pi. 6,a,b). The fibers diverge outward 

 from the center of the sclerodermites (pi. 6,c,d) in much the same 

 way the fibers of the coenenchymal spicules diverge outward from 

 the spicule core, and I suggest that the sclerodermites of Isis arei 

 actually highly modified spicules. This suggestion is corroborated by = 

 a comparison of the axis of Isis with that of Parisis, an undisputed 

 scleraxonian with spicules in the horny nodes. In that genus we see i 

 similar but shorter sclerodermites radiating out from the axis core; 

 the secondary thickening consists of sclerodermites oriented longitu- 

 dinally (pi. 4,6) and continuous with the spicules imbedded in the' 

 horny nodes (pi. 4,a). 



In the translucent, glassy axis of Primnoisis (pi. 6,e) the radial i 

 sclerodermites extend from the axis core to its surface as more or less 

 regular prismatic rays, just as in the Ellisellidae. In Acanella and 

 Keratoisis the orientation of calcareous elem.ents is clearly radial but 

 individual sclerodermites are obscured. 



Order PENNATULACEA Verrill, 1865 



Structure of the colonial axis : Since nothing of pennatulids but 

 the axial rod is ordinarily fossilized, intrinsic features of these struc- 

 tm'es must be sought for generic separation of fossil material. Exter-r 

 nal form and sculpture are too variable to be of use, so internal' 

 structure and system of calcification is the last resort. ^ Thin sections 

 of the axis show that its pattern of calcification (pi. 7) is strongly 

 suggestive of the Ellisellidae (as Kolliker pointed out) and Isididae: 

 Longer or shorter, continuous or interrupted prismatic^or u-regularly 



