104j me. p. h. caepentee on the genus actinometea. 



never seem to undergo calcification, though this may take place in Aiiledon Esch'lchtii, 

 either only very slightly, as in the two specimens examined hy myself (PL IV. fig. 10, »S'), 

 or to a considerable extent, as in that figured by Loven. In Actinometra also the extent 

 of calcification of the rays of the basal star is very variable. In one specimen of Act. 

 ])oIymorpha scarcely any trace of them is visible (PI. VI. fig. 11, S) ; in another they are 

 short, but otherwise well developed (fig. 1, S) ; while in others they may extend very 

 nearly to the outer angles of the radial pentagon (figs. 13, 18, 19, 22, 2Jf, S). 



In Act. pectlnata, again, we have found them to be large and tliick (PI. V. fig. 8, -S), 

 while in the closely allied Act. Solaris (PI. V. fig. 3, <S') they were slender rods, only imper- 

 fectly calcified here and there — the intervening portions of the dorsal interradial furrow 

 seen in the prepared skeleton (d./f) being occupied in the fresh state by masses of fibrous 

 tissue, which are removed by the action of the alkali used in preparation. 



(§ 71) We have seen that the basal circlet of Actinometra is somewhat complicated 

 in its nature, and consists of two entirely distinct elements, viz. a central rosette, which 

 we may fairly suppose to be the result of the metamorphosis of the embryonic basal 

 plates, as in Antedon rosacea, and five more or less completely ossified rays extending 

 from it in a peripheral direction. 



The rosette is situated on the ventral side of the quinquelocular organ, from the fibrous 

 envelope of which proceed the primary basal cords. These are very short, and soon 

 bifurcate, so as to give rise to ten secondary cords, which pass through the ten apertures 

 {vi, Vi, . . . Ci, Z2) in the peripheral portion of the rosette in order to reach the central 

 canals of the first radials. 



As already remarked by Ludwig , we may fairly regard those elements of the skeleton 

 in which the bifurcation of the primary basal cords occurs as homologous throughout 

 the ditl'erent ijeuera of the Criuoids. This leads us to the conclusion that the rosette 

 of Antedon and Actinometra is homologous with the united central ends (at least) of the 

 basals of Fentacrinus, which are perforated by canals that lodge the five bifurcating 

 fibrillar cords proceeding from the dorsal angles of the quinquelocular organ, and not 

 from its ventral angles as in Comatula. 



The question now arises, Where are we to seek for the homologues of the five rays of 

 the basal star in most Comatidcc? Unfortunately the only type in which we find 

 a condition any thing like that described above is a fossil one, the Solanocrinus of 

 Goldfuss ; so that it is difficult to ascertain the precise relation of its basals to the canals 

 proceeding from the quinquelocular organ that was undoubtedly contained in the cavity 

 of its deep centrodorsal piece. 



The upper surface of the latter, according to Goldfuss", presents "five radiating 

 elevations on Avhich the pelvis articulates." They correspond to the interradial elevations 

 on the ventral surface of the centrodorsal piece oi Antedon Vind Actinometra. 



The basals themselves vary in appearance in the dilferent species. In S. costatns and 

 aS'. scrobiculatus they are only " fiiuf schmale Strahlen die sich zwischen die Nahte der 

 Rippenglieder eiusenken;" but in aS*. Jcegeri they are nuich wider, " so dass sic auf der 

 ganzen Gelenktiache zusammenstossen und hier fimf ausstrahlende Purchcu zur Aufnahme 



' Beitriigo &c. he. cit. p. 07. ^ Petrcf. Germ. loc. cit. p. ICG. 



