TAXOCRINIDAE 427 



second or third IIBr. The series also shows a corresponding progression in the lateral 

 union of the pairs of IIBr forming the continuation of the main ray, from a partial attach- 

 ment of one pair in the young to a union of three directly, and of more by the aid of illBr in 

 the mature. 



Plate LXVII contains a series of preparations for the special illustration Of the teg- 

 minal, anal, and oral structures. Figure 1 gives a dorsal view of the anal tube and the 

 form of the posterior basal, as seen upon a specimen with rays horizontally extended. 

 Figure 2<z is a ventral view of a similar specimen with the tegmen preserved but somewhat 

 broken down, including the plated perisome, parts of ambulacra, oral plates and those con- 

 necting with them ; and especially the anal tube, which is quite perfect, curving to the right 

 behind the posterior oral, with the folded perisome attached for its full length at one side. 

 All of these appear more clearly in the enlarged figure of this central portion, 2b; the fold 

 in the perisome at the right of the tube shows with accuracy the relation of the tube plates 

 to the perisome, and that when in life this articulated series of strong plates was erected 

 it simply raised a portion of the flexible integument into a conical tube, forming the rectum, 

 to a height where the excrement could be discharged away from the oral center. 



In figure 3a the oral structures have fallen to the rear, partly overlying the tube with 

 the perisome still attached at both sides ; and 3b, an enlarged view from .a more favorable 

 angle, shows the tube to its distal end as if it were imbedded in the perisome ; the aperture 

 cannot be identified, being doubtless always closed by pressure upon the perisome. Figure 4 

 is a detail from a large specimen in which the exact mode of connection of the perisome to 

 the margin of the anal tube is shown ; it is attached on both sides in its original position 

 for a distance of four or five plates ; at the fifth the tube has fallen over nearly at right 

 angles, and the upper part has slipped away from the perisome, leaving exposed the grooved 

 margin by which it was attached. In figure 5 we have the inner side of a tube from near 

 the distal end, with the tubular wall of perisome attached but flattened by pressure ; the cross- 

 section shows how it was attached to both margins of the anal tube plates. This figure 

 leaves no doubt that the series of strong, articulated plates is really the backing and support 

 of the rectum formed by a conical protrusion of perisome, thus constituting the anal tube of 

 which these plates are the outer wall. Figure 6 is a fragment of similar tube plates of 

 0. exsculptus, with the marginal crease for attachment of perisome well exposed. The tube 

 is always found curved to the right under the edge of the posterior oral or its supporting 

 plates. 



Passing now to the oral structures, we have in figure 7 an almost complete tegmen, with 

 very little disturbance of the parts out.of their natural positions. The tegmen of the Flexi- 

 bilia is so fragile that in most cases where the ventral side can be got at in the specimens the 

 parts are broken down and piled in a confused mass at the bottom of the calyx. Such was 

 the case with several of this species used during the earlier searches of Wachsmuth and 

 myself for the ventral structures ; after patiently excavating the matrix with which every one 

 of these cavities was solidly filled, we would come to a large, uncouth looking object at 

 the very bottom having no apparent relation to anything around it. Nothing more being 

 seen, we would give that up as a failure and try another, only to be followed by many such 

 disappointments. That large, uninteresting object, which I supposed to be some foreign 

 body accidentally lodged there when the crinoid was imbedded, was the posterior oral 

 entirely displaced, with the other parts disintegrated beyond recognition ; but I did not know 

 it then, nor for long afterwards. 



It was many years after the acquisition of these specimens, and after the fruitless 

 efforts above described to find the ventral structure, that I again took up the search in the 

 Indian Creek specimens. They were the most encouraging because of the firm and even 

 texture of the matrix, and its freedom from pyritous nodules and incrustations which destroy 

 the finer structures in the cavities of the Crawfordsville specimens. A number of them 

 28 



