I36 SMITHSONIAN INSTITUTION 



variations in size from the smallest, 3.5 mm. in diameter, to the largest, 11 mm. A large part 

 of the spheroid is taken up by the radials, which occupy over two-thirds the surface of the 

 calyx, while the basals and radianal are small. The anal plate thus becomes proportionally 

 long and narrow ; it only projects about one-third its length above the radials, and the pro- 

 jection is shorter and more obtuse in young specimens. The primibrachs are constant at two 

 in 60 rays of 19 specimens having more or less of the arms preserved. 



By the above proportions of the calyx plates, as well as by the thinness of the walls, it is 

 well distinguished from L. pusillus of the earlier shales. By comparing the numerous figures 

 on the same plate, it will be seen how much larger are basals and radianal in that species, 

 tending to encroach upon the radial circlet, giving it greater width and less proportionate 

 height ; so that the anal plate, although projecting for nearly half its length above the calyx 

 level, is nevertheless but little if any longer than wide. There is also a constant difference 

 in the form of the column facet, which in this species is sharply excavated and indented, 

 leaving vertical sides, whereas in L. pusillus it is shallow and saucer-shaped ; and it is accom- 

 panied by more protuberant infrabasals. The stem of the two species is decidedly different, 

 that of L, pisiformis being considerably unlike the prevailing form in the genus, and exhibit- 

 ing in its very long columnals of uniform width without intercalation of short ones between 

 another juvenile character (PL I, fig. 16). On an average L. pusillus is much the larger 

 species of the two, although specimens of it are found which are smaller than the maximum 

 size of L. pisiformis. It has also much stronger and thicker calyx walls (PL I, figs. 5, 18). 



There is no other described American species with which this one need be compared 

 except L. hemisphericus Rowley, which is evidently founded upon a very young specimen of 

 this species. I have figured Professor Rowley's types (PL I, figs. 31, 32), and it would be 

 difficult to separate them from many of the Tennessee specimens of like size and proportions. 

 I have also figured two specimens from Jefferson County, Kentucky (figs. 34, 35), which I 

 refer to this species with some doubt. They were in the Lyon collection, found by Col. Lyon 

 in the Niagaran limestone near Louisville, at a horizon considerably higher than that of the 

 Tennessee beds. These two specimens are of maximum size for the species, figure 34 being 

 larger than any from Tennessee ; they are somewhat flattened, but appear proportionally 

 wider, and the anal plate broader, than is usual in the species. The strongly beveled edges 

 of the brachials indicate a mature growth, but some approach to this can be seen in other 

 specimens. Both specimens have a little different aspect from those from Tennessee, and 

 the basals are proportionally larger ; but they have the same infrabasals, indented column 

 facet, and juvenile proportions of the arms. I am unable to point out constant differences on 

 which to base a new species for them, and I think they must be regarded as representing an 

 advanced stage of growth in the present species. 



L. pisiformis is more like the Gotland species L. facietatits than any other, and I was at 

 first inclined to consider them identical. They have the same thin calyx walls, large radials, 

 and long, narrow anal plate, projecting but little above the calyx; but on comparison the 

 Swedish form has a broader and more truncate base with more prominent infrabasals, and 

 the anal plate is much less angular above. 



In this prolific species, as might be expected, some abnormal examples are to be found, 

 and there would doubtless be more but for the fact that in most of the unweathered specimens 

 found in the colony near Decaturville, Tennessee, the sutures are difficult to see. In one the 

 radianal has almost disappeared, only the smallest triangular vestige of it remaining (PL I, 

 fig. 27) ; and two others (figs. 28, 29) have no radianal; these are of especial interest in con- 

 nection with the literary history of the species. 



L. pisiformis has been buffeted about considerably in the books. Meek and Worthen 

 very properly demurred to Roemer's reference of it to Pot eriocr inns, because it has only one 

 anal plate, and they suggested that it might belong to their genus Arachnocrinus, then known 

 only from Hall's Cyathocrinus bulbosus. Hall in the 20th Regent's Report mentioned 



