NO. 3 ON THE FOSSIL CRINOID FAMILY CATILLOCRINIDAE 9 



almost the whole of the pelvis." He recognized only two series of 

 calyx plates, and not three as Shumard did ; hence the " pelvis " of 

 Troost is the same element that is called " primary radials " by 

 Shumard. 



Thus, while the generic diagnosis shows a divided base of three 

 plates, the specific description, like Shumard's, says nothing about 

 their number. Miss Wood (Op. cit., p. 24), alluding to this, and to 

 the statement by Wachsmuth and Springer (Revision III, p. 368), 

 that the base is undivided, says : " Troost's observation appears to 

 be correct, as the suture lines between the plates show distinctly on 

 one of his specimens, and traces of them appear upon others." 



I have the Troost types before me ; the original of Miss Wood's 

 figure 3 of plate 9 shows one very distinct suture under r. post. R, 

 and two others, faint but distinguishable in a proper Hght, under 1. 

 post, and r. ant. RR. In the manuscript of Troost's monograph there 

 is a diagram of the cup (not reproduced in the publication) showing 

 two distinct interbasal sutures, and a third one indicated by a dotted 

 line. The three sutures do not meet at the axial center, but connect 

 with a small obtusely pentagonal area in which the thin plates are 

 broken out. 



Thus the actual construction of the basal ring has been a matter 

 of doubt among different observers, Troost believing it to be com- 

 posed of three plates, and Wachsmuth and Springer, and others fol- 

 lowing them, considering it as undivided ; while Shumard puts the 

 number as doubtfully five, with another series below or within them. 

 It is important to ascertain, if possible, what the fact is. 



COMPOSITION OF THE BASE 



In addition to the Troost types in the original Troost collection in 

 the National Museum, I have a number of good specimens of the type 

 species from the two localities at which it has been chiefly found. 

 Button Mould Knob, Kentucky, and White's Creek Springs, Tennes- 

 see ; most of them consisting of the dorsal cup only. Careful exami- 

 nation under the best light conditions, including cutting and polishing 

 of some specimens, fails to disclose distinct, unmodified interbasal 

 sutures ; but in all the lines of division between the plates are more 

 or less obscured, either by compression, by secondary growth of 

 stereom, or by molecular changes during the process of fossilization. 

 The calcite of which the specimens are composed when not silicified 

 is usually crystallized, developing its characteristic lines of cleavage 

 which tend to obliterate fine sutures — even the basiradial sutures. 



