LECANOCRINIDAE 195 



proportions of the plates." The average size of the specimens shows a constant difference, 

 and the column, as shown by the British Museum specimen, is so different that in a less 

 specialized form it would indicate a different genus. That of M. forbesiocrinus not only 

 lacks the spindle-shaped median columnals, but is a perfectly even, straight-sided cylinder, 

 without any depression at the sutures or notable increase in length of columnals. 



As to Hall's other species, M. scitulus, I think there can be no question but that it is iden- 

 tical with this. It comes from the same bed in the Lower Burlington limestone at Burlington, 

 and as a matter of practice I was never able to recognize any material difference among the 

 specimens until I resurrected the type specimen of M. scitulus in the Museum of Comparative 

 Zoology, where its identity had been lost (PI. V, figs. 14a, b). Hall differentiated the second 

 species on the greater angular prominence of the " basals " (IBB) — which were scarcely 

 visible in his very small specimen of M. konincki, though well developed in others — and in 

 the more rapid spreading of the calyx, which in his type specimen is due to foreshortening by 

 vertical pressure. The difference in the infrabasals is, however, very remarkable, those of 

 M. scitulus being not only more angular, but being distinctly five in number — a departure 

 from the structure of the genus and of the group generally. Nevertheless I do not consider 

 it in this case to be even a specific character. 



For some unaccountable reason there is in this highly specialized form, with its dex- 

 trorse twist of arms and its strange column, a tendency in the infrabasal ring to revert to the 

 fundamental five plates by division of the two larger ones. This happens in an irregular 

 way. In the type of M. scitulus the plates are of unequal size. In Plate V, figure 11, the 

 regularly shaped infrabasal disk is divided into five equal plates, while in another specimen 

 the tendency to such division is to be seen, not developed into actual sutures, but visible by 

 transmitted light in the semi-translucent calcite of the plates. In the specimens described by 

 Miller and Gurley as Cyathocrinus blairi there are said to be five infrabasals, and in a speci- 

 men labeled as that species from the same locality, in the Gurley collection at the University 

 of Chicago, only three are visible (PI. V, fig. 18). This is simply a case where under the 

 influence of strong specialization the bond even of ordinal characteristics becomes weakened, 

 and the tendency to sporadic variation is increased. Miller, after describing his specimens as 

 Cyathocrinus, noted the structural similarity to the Ichthyocrinidae, and afterwards dis- 

 covered their apparent resemblance to Mespilocrinus; but on account of the difference in 

 number of infrabasals he denied that they could belong to that genus, though in this case he 

 considerately refrained from proposing a new genus for them. To be consistent, he would 

 have had to establish a new family for the reception of certain specimens of a form which he 

 could not have distinguished, even specifically, if the infrabasals were concealed — thus show- 

 ing the absolute lack of that correlation of characters which is necessary to the validity of 

 such divisions. 



Type. The location of the type specimen is unknown. The figure is made from 

 Professor Whitfield's drawing for Hall's photographic plates, together with a cast of the 

 type made by Whitfield while it was in Hall's hands for description ; this cast is in the 

 American Museum of Natural History, New York. 



Horizon and locality. Lower Carboniferous, Lower Burlington limestone and ranging 

 rarely into the Upper Burlington ; Burlington, Iowa. 



