Vol. 54.] MR. P. A. BATHER ON PETALOCRINUS. 415 



in the much older specimen e ; while the width of the ridges and 

 grooves in g is as great as, or greater than, in f and e (PI. XXY, 

 figs. 1, 4, & 6). The difference in size that does come about is in 

 the width of the facet, and to a less extent in the width of the 

 grooves of the I and II Br. This, however, is produced purely by 

 lateral accretion : there is no intussusception, nor even any absorp- 

 tion of stereom ; as the facet widens, so also the outer ridges 

 increase in width and height at their proximal ends. 



The narrower ridges appear to have rounded upper surfaces ; but 

 the broader ridges are flat or slightly concave, and their edges 

 appear sharp and square. The variation in the ridges, as between 

 species, will be more fittingly discussed under the heads of the 

 species themselves. 



Mention has just been made of covering-plates (ambulacrals). 

 None have yet been found in any of the Gotland species. It is just 

 conceivable that a minute mass, apparently of stereom, that seems 

 to roof over a little piece of one groove in P. longus may represent 

 one or two such plates (see p. 433). They certainly are not preserved 

 in any other of the American specimens. Prequeatly, however, one 

 can distinguish, along the sides of the grooves, notches apparently for 

 their reception. The following description applies to P. vishi/censis, in 

 which the notches were first observed and studied. In the flattened 

 ridges they are not visible from above, but when the ridges are 

 viewed from the side, with a strong light directed down the grooves, 

 then the notches can be seen immediately under the straight edge 

 of the ridge. Of these notches 4| go to 1 mm. in specimen p 

 (PI. XXV, fig. 25). In the forms with narrower rounded ridges 

 the notches are visible from above, since they lie on the edges of the 

 ridges. In specimen e about 4 notches go to 1 mm. (PI. XXV, 

 fig. 24). The notches on one side of a ridge alternate with those 

 on the other side, and it appears also that the notches on one side 

 of a groove alternate with those on the other side. This alternating 

 arrangement would be a natural outcome of the alternation of 

 the covering-plates. We infer, then, that there were ambulacrals 

 alternating as usual, with a width of '25 or -22 mm., and some- 

 what closely articulated to the sides of the arm-groove. Their 

 absence in the case of arms that could not be folded up, that 

 became separated from the calyx after death, and that lay about on 

 the sea-floor, is nothing more than is usual. 



Notches for covering-plates are also seen in P. aiigustus, where 

 they are very clear (PI. XXV, figs. 26 & 27), in P. mirabilis, where 

 they are usually obscure (Pi. XXVI, fig. 55), and in P. longus, 

 where they are seen most clearly on the thin regions of the ridges, 

 and give them a saw-like appearance (PI. XXVI, figs. 61 & Q2). A 

 rabbet or ledge, apparently for the support of the covering-plates, 

 is occasionally to be distinguished a little below the ventral edge 

 of a ridge ; as, for example, in P. longus (PL XXVI, fig. 62). 



In P. mirabilis fl, the arm numbered ii has lost, in its right- 

 hand half, a portion of the fioor of the grooves, thus exposing what 

 lay on the ventral surface of the arm. Here are seen some eighteen 



Q. J. G. S. No. 215. 2 G 



