414 MR, p. A. BATHEE ON PETALOCEINUS. [Aug, 1 898, 



distal ends of the grooves by weathering, a process that destroys the 

 indentation of the ridges. This indentation is continuous with the 

 depressions on the dorsal surface, and also passes on to the distal 

 ends of the ventral ridges, even when the ridges themselves are 

 narrow and rounded for the greater part of their course. A well- 

 preserved distal margin of an arm-fan thus reminds one of the edge 

 of a roof formed of concave tiles, \J'^^j\j\j\j . When the ridge is 

 broad, the depression often passes for some distance along its ventral 

 surface, making it slightly concave. 



There is considerable variation in the width of the ventral ridges, 

 both absolutely, and relatively to the grooves. Thus in P. longus 

 some ridges are almost knife-edged (or, as will be seen, it would be 

 better to say ' saw-edged'), having in parts a width of '2 mm.; while 

 in P. eocjpansus a width of 1*9 mm., nearly 10 times as much, has 

 been noted. There is variation, not only between species, but 

 between individuals of a species, between ridges of an individual, 

 and between different regions of the same ridge. This is exemplified 

 by the following study of P. vishycensis : — First, in each individual, 

 while the grooves remain of approximately the same width, only 

 tapering gradually distalwards, the ridges widen distalwards up 

 to each branching, when they rapidly become narrow again : this 

 is well shown in g and h (PI. XXY, figs. 1 & 8). Again, in some 

 individuals the ridges are far wider than in others. Thus in j 

 the ridges are quite sharp in appearance, varying in width from 

 •2 to '6 mm., while the normal width of the grooves tapers 

 from about -7 to '5 mm. In e the ridges are of aboat the same 

 width as in j, but they are not so high, and the grooves are not quite 

 so wide, so that the ridges do not seem so sharp (PI. XXY, fig. 6). 

 In c, the width of the ridge varies from -4 to 1 mm., while the 

 normal width of the grooves is about '5 mm. In p, the width of 

 the ridges varies from -5 to "9 mm., and the normal width of the 

 grooves is also '9 mm. (PI. XXV, fig. 9). The greatest width of a 

 ridge found is in the fragment lettered m, where the width of the 

 ridges between the outer IVBr on each side attains I'o mm. ; the 

 width of the grooves at this level is '75 mm. 



Since the sides of the grooves slope inwards, it might be suggested 

 that this variation in the width of the ridges of different specimens 

 was due to the greater or less wearing down of the specimens ; but 

 this is negatived by the preservation of notches for the covering- 

 plates, as described in the sequel. It might also be thought that 

 some variation was due to the dififerent ages of the specimens : if 

 the arms were not fused together, then the sides of the proximal 

 brachials would undoubtedly grow wider at the same time as the 

 whole arm increased in length ; and the corresponding mode of 

 growth in Petalocrinus would be an increase in the width of the 

 ridges by intussusception. A comparison of young specimens with 

 old does not, however, lend colour to this supposition : the ridges 

 and grooves at their first appearance have a width similar to that 

 obtaining in more developed specimens. Por instance, in the young 

 specimen a, the width of a ridge is actually half as great again as 



