Vol. 54.] ME. F. A. BATHER OX PETALOCRINTTS. 419 



places, distinguish the dark lines caused by axial canals, which were 

 ground through from the dorsal side before the ventral grooves 

 themselves were reached. Sometimes, however, towards the distal 

 margin, it appears that the canals were not completely separated 

 from the grooves, this separation being, as is usual among Inadunata, 

 a subsequent process caused by stereom growing in from either 

 side and meeting in the middle so as to shut off the axial nerve 

 from the other organs contained in the ventral groove. This is 

 prettily shown by the broken surface of specimen p (PI. XXV, 

 fig. 10). Here the canals on the right are cut across more proxi- 

 mallj than those on the left. Canal i is completely closed in ; ii 

 is also closed, in, but there can be seen at the bottom of the ventral 

 groove a little notch indicating the line along which the two halves 

 of the stereom-bridge have met ; iii is less far advanced, in that 

 the stereom-bridge is only about halfway across ; while in iv the 

 axial canal is only distinguishable as a narrower U-shaped tongue 

 at the bottom of the ventral groove. These stages may be compared 

 with various figures of brachials of Gissocrinus given in ' The Cri- 

 noidea of Gotland, I,' as, for example, stage ii, with figs. 289, 299, 

 338, & 358 ; stage iii, with figs. 270 & 333 ; stage iv, with figs. 300, 

 311, 319, 332, & 366. Many specimens show a line along the floor 

 of the grooves, caused by the imperfect union of the two sides 

 of the stereom-bridge. On the other hand, the bridge sometimes 

 appears to rise a little along the middle line, so that in the 

 horizontal section a thin strip of the stereom (6 in text-fig, 7, 

 p. 418) appears to lie in the middle of the ventral groove. 



These conclusions are confirmed by specimens of other species. 

 In the type of F. angustus the proximal end is broken and shows 

 two axial canals, which, however, are rather wide, since they are 

 just branching into four. In P. mirabilis, specimen d 2, the distal 

 ends of the grooves have a section like that of groove iv in PL XXV, 

 fig. 10 (PI. XXVI. fig. 44). The type of P. longus shows the canals 

 coming to light on the distal margin, the line on the floor of the 

 ventral groove, and the breakings through the floor of the groove 

 (PI. XXVI, fig. 64). The axial canals were not mentioned in the 

 original description of the genus. 



In describing the facet of P. vishycensis^ specimen i, mention was 

 made of a sharp median vertical depression, continued on to the 

 median ventral ridge. The meaning of this may be gathered from 

 the horizontal section of w (text-fig. 7). In this are plainly visible 

 fine dark lines running along the middle of the ridges. These 

 doubtless represent the original spaces between the primitively 

 independent arm-branches ; in other words, the sutures along which 

 they fused. The suture (s) is more pronounced in the median ridge, 

 that is, between the two main divisions of the arm. Here also one 

 sees evident traces of it in P. visbycensis, specimen n ; and consecu- 

 tive transverse sections showed its gradual commencement and 

 increase in size distal wards (text-fig. 6). In P. angustus the ridges 

 are marked along their middle lines by dots of black pigment, 

 .which appear to be traces of the original connective tissue along the 



