URASTERELLHLE. 131 



many of the apical plates to be "drawn out into more or less long, blunt, stout, 

 erect, non-articulatory rods." Hudson has shown more recently that the rods 

 carry small spines, and the plates are in fact paxillae. It is convenient for descriptive 

 purposes to adopt Ludwig's terms (40, p. 510) for the various parts of the paxilla, 

 namely, paxilla-base, paxilla-shaft, and paxilla-crown. These terms are illustrated 

 in Text-fig. 82. 



One of Hudson's figures is reproduced here (PI. VIII, fig. 7). Each paxilla 

 according to Hudson, carries " three long articulated spinelets " (93, p. 130). He 

 also shows that the paxilla-shafts (pedicels) and spinelets " are built of an alter- 

 nating series of light and dark discs. The dark discs indicate the former presence 

 of organic tissues, the white discs the presence of more open or spongy stereom 

 formation. The writer would interpret this appearance as indicating that the 

 spinelets were increased in length by a series of tissue extensions at the tips, these 

 extensions becoming consecutive centres of stereom formation which, however, did 

 not completely join one another. The spines seem also to have been formed in the 

 same manner, and the whole structure is of a very primitive nature. Such a 

 structure must not only have kept the spinelets from becoming very rigid bodies, 

 but it must also have allowed them to fall apart, like a broken string of beads in 

 decay . . . the pedicels also became separated from the plate-bases during 

 decay and, like the spinelets, broke up into similar but more robust beads" 

 (op. cit., p. 131). 



I have not been able to distinguish the spinelets in English forms, possibly 

 because they so readily decay. The paxilla-shafts (pedicels) are, however, usually 

 recognisable. In most species of Vrasterella (Text-fig. 92, p. 143) the paxilla- 

 shafts of the radialia and the first row of adradialia are ridge-like, the ridge 

 running along the length of the ossicle, while the remaining adradialia have rod- 

 like paxilla-shafts. In Salteraster (Text-fig. 94, p. 150) the paxilla-shafts of the 

 radialia are usually boss-shaped, those of the adradialia rod-like. In Urasterella 

 montana they are shaped like an inverted cone (Text-fig. 93, p. 148). 



Structure of the Ambulacral Groove (Plate VIII, figs. 1 — 3; Text-figs. 83 — 85). 

 — Many mis-statements have been made with respect to the structure of the ambu- 

 lacral groove. Most of them have been corrected by Hudson (op. cit.), but it is 

 still possible to add details to the facts already established. Some of these details 

 allow one to obtain a more adequate picture of the mode of life of the forms. The 

 material I have studied the most closely is that of Vrasterella thraivensis, n. sp., 

 several of the specimens of this species showing displaced isolated ossicles. It was 

 found possible to make enlarged models of them in plasticine, which were later 

 reproduced in plaster-of-Paris. A photograph of the more proximal portion of 

 the ambulacral groove reconstructed in this way is given (PI. VIII, fig. 1). The 

 position of the ossicles is almost that which they would occupy if the groove 

 were widely open. A lettered tracing of this photograph is given (Text-fig. 83). 



