36 THE VOYAGE OF H.M.S. CHALLENGER. 



there is a complete ring of united basals as in the recent genera Atelecrinus and 

 Thaumatocrinus (PL LVI. figs. 1-4). In the four other genera of recent Comatulidse, 

 however, the primary embryonic basals undergo transformation into the well-known 

 rosette, which is really a secondary structure. It lies over the chambered organ, between 

 the under surface of the radial pentagon and the^ upper face of the centro-dorsal ; but 

 it is entirely concealed, and does not appear at all upon the exterior of the calyx. 



In many Comatula^ more or less prismatic rods proceed outwards from the inter- 

 radial angles of the rosette, and their ends are often visible on the exterior of the 

 calyx. They occupy exactly the same positions as the rod-like basals of many fossil 

 Comatulse, but do not represent them morphologically, for they are not developed from 

 the primary embryonic basals. These last become transformed into the rosette, which 

 is a secondary structure as shown by Dr. Carpenter ; ' while the tertiary basals which 

 are connected with its interradial angles are merely ossifications in the connective 

 tissue of the synostosis between the radials and the centro-dorsal, and are somewhat 

 variable in their development.^ They do not occur in Antedon rosacea, nor in the 

 species on both sides of the Atlantic which are most nearly allied to it, viz., Antedon 

 phakmgium, Antedon dentata, Antedon hageni, &c. ; and they seem to be absent in 

 Eiidiocrinus, at any rate in Eudiocrinus semper i. But they reach a relatively large 

 size in many tropical species both of Antedon and of Actinometra, in which latter 

 genus I have never found them to be absent, and they also occur in Promachocrinus 

 (see fig. 1, A on p. 37). [See Note B.] 



B. The Eadials. 



There is considerable variation in the degree of lateral union between the individual 

 basals and radials of stalked Crinoids. Those of Holopus (Pis. I. -IV.) are so intimately 

 fused that the sutures are entirely invisible on the exterior of the tubular calyx. In 

 Hyocrinus (PI. VI.) the interradial sutures are quite evident, but those between the 

 basals are very obscure. In Rhizocrinus, and especially in Rhizocrinus Iqfotensis, the 

 sutures between the basals are usually entirely invisible, not only externally but also 

 in transverse sections of the decalcified calyx ; while the basiradial and interradial 

 sutures are merely indicated by very faint lines on the surface of the cup (PI. IX. 

 figs. 1,2); and strong measures are necessary before the individual joints will se23arate 

 from one another. In the Pentacrinidae and Comatulidas the union between the radials 

 is less close and the sutures well defined, while the condition of the basals varies accord- 

 ing to circumstances (PI. XIII. fig. 1 ; PL XV. figs. 1, 2 ; PL XVI. ; PL XX. ; PL XXXIII. 

 figs. 8-10 ; Pis. XXXV.-XXXVII. ; PI XXXIX. ; PL XLIIL). The union of the 



' Phil. Trans., 1866, pp. 744, 745. 



'' The genus Actinometra, Trans. Linn. Soc. Lond. (Zool.), vol. ii. pp. 93-105 ; and the genus Solanocriuus, Journ. 

 Linn. Soc. Lond. (Zool.), vol. xv., 1880, pp. 212-214. 



