Vol. 54.] MR. P. A. BATHER ON PETALOCRINTJS. 421 



no great error, since experience teaches that in highly-modified 

 genera or families (as, lor example, H&rpetocrinus, Caiceocrinidse, 

 Eucalyptocrinidse), the structures in which greatest differences 

 between species are manifest are those in which the principal modi- 

 fications characterizing the genera have taken place. This statement 

 does not apply (under classifications based on the premisses usually 

 accepted) to genera in the beaten track, that do not diverge widely 

 from the common path (such as Cyathocrinus, Pentaorinus, and 

 Platycrinus), for in them the criteria are usually small differences 

 of ornament or of measurement, to human intelligence often 

 unimportant, and stigmatized as ' useless characters.' 



tSpecies of this latter type owe their differentiating features, in 

 many cases, to difference of ancestry. They go in crowds, each 

 heterogeneous in origin, but moulded to a generic pattern by common 

 conditions, and passing sheep-like from the fold of one genus into 

 that of another. Species of the former type belong to a vigorous 

 and sportive strain, started in a few individuals, and speedily over- 

 leaping the barriers of the flock ; the essential features of the 

 ancestor are preserved, and the race, after running rapidly through 

 its extravagant changes, becomes extinct, as though its energy were 

 spent. The variations produced have been of no advantage ; they 

 have broken against, rather than been shaped by, the inevitable 

 shears of selective circumstance. 



Application of the foregoing to the aberrant PetaJocrinus suggests 

 that the arms, as they are the structures most divergent from the 

 normal crinoid type, will also be those in which variations suitable 

 for discrimination of species are likeliest to occur. 



Por the present, ordinal and family characters need not be dis- 

 cussed, but the differentiating features of Petalocrinus, described at 

 length in the preceding pages, may be formulated in the following — 



Generic Diagnosis. — Base dicyclic ; IBB minute (probably 

 fused to 3) ; BB 5 ; RR 5. Arms inadunate, distinct ; each com- 

 posed of IBr^, united to E, by perforate articulation, and an arm-fan, 

 similarly united to IBr^ and formed by the anchylosis of all dorsal 

 elements of a non-pinnulate dichotomous arm from lax onward. 

 Tegmen solid, containing 5 large iAmb ( = deltoids or orals). Stem 

 subcircular m section. 



Genotype : P. mirahilis. 



P. visbycensis, n. sp. 

 (PI. XXY, figs. 1-25.) 



Angle of arm-fan 70° to 93°, usually 80° to 85° ; fan bilaterally sym- 

 metrical in shape, but not quite so in arrangement of grooves ; finials 

 10 to 22; ventral surface of fan gently convex ; dorsal surface slightly 

 concave, with growth-lines and shagreen ornament ; ridges usually 

 as wide as, or rather less wide than the grooves, often flat-topped. 



Type : Specimen e, Eiksmuseum, Stockholm (PL XXV, figs. 5, (j, 

 7, 17, 23 & 24). Paratypes in Riksmuseum and British Museum. 



Locality : Shore north of Wisby (Gotland). 



Horizon : Lower Silurian (as explained on p. 403). Beds b & 

 C of Lindstrom. 



