f a i. \ mock i xrs mo.Mr.n.K. 



compact reticular system to the delicate open system of meshwork imme- 

 diately surrounding the opening of the axial canal (Plate XXI. Fig. I). 

 The lower face of the upper joint (Plate XXI. Fig. 2) is flat, and the 



raised angular ridges are less marked, but Otherwise it does not differ 

 materially in its structure from that of the upper face. The quinque- 

 lobular disk (Plate XXI. Fig. 3) lias the same reticulation with the inner 



par! of the stem joints. It has a delicate open meshwork surrounding the 

 axial canal, which passes gradually into a closer and more compact lime- 

 stone plating towards the periphery and in the wedges forming the angles 

 between the adjoining indistinct lolies. This disk seems to differ in its 

 mode of origin from that of new joints, and 1 am inclined to look upon 

 this disk as the representative of the anchvlosed infrabasals detected in 

 the Apiocrinida. 1 .* The new joints, if we can judge from the figures 1 

 have given, appear at first towards the outer edges of the older rings as 

 delicate limestone spurs extending towards the centre, while in the case 



* With the exception of Encrinus and Apiocrinus, the intranasals have not been described in the 

 more recent types of Crinoids, but they arc known in about half of the palaeozoic Crinoids. 



The infrabasals of Encrinus, as figured bj Beyrich (Crinoiden d. Muschelkalks, p. 1, Plate I. 

 Kig. l c ), are entirely enclosed and hidden by the upper stem joint, .and occupy exactly the same posi- 

 tion as the lobed plate I have described at the summit of the uppermost joint of Calamocrinus. 

 l)e Loriol (Monog. des Crinoides de la Suisse, p. 7. 1378-79) considers the supplementary basals 

 of Encrinus as forming no part of the calyx, but as belonging to the uppermost stem joint, which 

 he has called tin- "article basal." De Loriol figures in the Paleontol. francaise (Tom. XI. 



1>- Parti.', i nl. . 1S82-S4) five small plates radially placed upon the centre of the summit of the 



t..p stem joint of two species of Millericrinus, M. polydactylus (Plate 110, Figs. I s , 2 a , p. 553) and 

 M. D'Orbignyi (Plat.' 111!. Figs. l b , l'\ p. 566). They do not appear at all upon tin' exterior of the 

 calyx, but rest upon the central pari of the enlarged uppermost stem joint, and are concealed by 

 the basals outside of them. 



Judging from the little we know of th.- young stages of the dicyclic palaeozoic forms as described 

 by Wachsmuth and Springer (Transition Forms in Crinoids, Proc. Phil. Acad., 1S78, p. 229), the 

 infrabasals are very highly developed, and gradually become less prominent as the young grow older. 



This seems to show that the infrabasals may have become rudimentary, or have been lost alto- 

 gether in those forms in which it has not been observed. The discovery of infrabasals by Bury in 

 ( lomatula ( Morphol. Studies, p. 1 is), and their subsequent resorption, goes far towards 

 making this a plausible explanation of the existence of infrabasals in the older types and their absence 

 in the more recent ones; although, as Neumayer justly remarks, we should be very cautious in taking 



rj| of (' alula, a geologically recent and quite aberrant type, as characteristic of that 



of the Crinoids as a whole. 



The infrabasals of Encrinus. which support the basals, are in Stemmatocrinus represented by a 

 Hit plate. Carpenter is inclined to look upon this as the top stem joint, the articlt l>,i«il of Apiocrinus 

 and .Millericrinus. This Hat. plate may have arisen from the anchylosis of the infrabasals, or it may 

 in some cases corresp 1 to the upper stem joint. This is the case in Rhipidocrinus, where the infra- 

 basals a:.- appar-ntly anchylo I it i ' central plate. 



II,. formation of the centrodorsal in Comatula and in M. Pratti, in which the infrabasals are 

 anchylosed with on o tern joints, may occur in other eases. 



But as we know the development of Crinoids only from a single type, the representative of a group 

 with a modified calyx, we should be careful, as Carpenter lias already suggested, in making com pari - 

 sons based upon einbryological data with the endless and complicated types of the palaeozoic genera. 



