A MONOGRAPH OF THE EXISTING CRINOIDS 327 



Hartlaub said that Carpenter evidently had several specimens from the Blake 

 collection that he referred to granulifera besides Pourtales' types, and he believed that 

 he considered as representing granulifera the specimens that he (Hartlaub) described 

 under that name, although no label in Carpenter's hand was found anywhere. 



Hartlaub remarked that his material showed to an extraordinary degree the great 

 variability of this form. He said that this variability is not confined alone to the out- 

 ward habitus or to the ornamentation, but affects also those features which ordinarily 

 show greater constancy, such as the distribution of the syzygies, the length of the 

 proximal pinnules, the relative visibility of the radials, and so forth, and in these cases 

 it is so extensive that at first it appeared to be impossible to give a specific diagnosis 

 that he could consider as of real value and reliability on the basis of the specimens 

 before him. Hartlaub considered it most practicable to give a detailed description of 

 the individuals from which an appreciation of the true specific characters may be 

 obtained. 



Hartlaub noted that in a letter to him dated December, 1909, 1 had sent him some 

 critical remarks on the proofs of his plates 1-10, and he found that I had described as 

 new species several of the forms regarded by him as varieties of granulifera. Thus the 

 second individual from Martinique as shown on plate 5, figure 8, I identified as my 

 Crinometra imbricata, and the first specimen from Martinique as shown on plate 5, 

 figure 3, I identified with my C. pulchra. Hartlaub said he need only remark that he 

 held to his original opinion and hoped especially that his plate 12, which I did not see, 

 would serve to justify it. 



In the two specimens from Blake station 45, which were part of Pourtales' original 

 material, the centrodorsal is dome-shaped, the bare dorsal pole with traces of cirrus 

 sockets. At the base of the centrodorsal are five small elevated interradial knobs. 



The cirri are about XX, 15-20, from 8 to 23 mm. long. The three basal segments 

 are thick, short, and cylindrical, those following elongate, the outermost shorter again 

 and strongly compressed laterally. There are no dorsal spines, with the exception of 

 the feebly developed opposing spine, but the distal ends of the segments are slightly 

 overlapping. 



The radials are short, though entirely visible, somewhat longer laterally than in 

 the median line, with a small knob in the middle of the dorsal surface; they are in close 

 contact laterally. The IBr t are somewhat longer than the radials; there is no note- 

 worthy elevation on the articulation with the IBr 2 (axillaries) ; they are laterally in 

 contact. The IBr 2 (axillaries) are pentagonal, laterally as long as the IBr,. The IIBr 

 series are either 2 or 4(3+4). In the smaller specimen there are three IIBr 2 and two 

 IIBr 4(3+4) series, in the larger five IIBr 4(3+4) series and one IIBr 2 series. In the 

 larger specimen the ossicles as far as the IIBr 2 have rather sharply flattened sides; in 

 the smaller this feature at the best is only suggested. The IIIBr series are 2, though 

 according to Carpenter they are 2(1+2); Hartlaub said that following a IIBr 2 series 

 the two elements of the IIIBr series gave the impression of being united by muscular 

 articulation. The surface of all the ossicles of the division series in the larger specimen 

 is flat and coarsely granulated, and in the middle of the dorsal surface there is dis- 

 tinguishable a stouter granule resembling that on the radials. Tins alone is present 

 on the smaller specimen, the rest of the ossicles being smooth. There are no synarthrial 

 tubercules. 



