114 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



marking the center of the resulting spiral, and the mouth lying above the outer- 

 most coil (see fig. 21, p. 69). 



The artieuliUion hy which the first post-radial ossicle is joined to the radial varies 

 greatly in the dill'erent groups (see figs. 9-11, p. 65, 31, p. 71, and 431, 432, p. 349). 

 The dorsal (outer) portion is occupied by a large, more or less semicircular or elUp- 

 soid depression, the Dorsal ligament fossa which is bounded vcntrally (anteriorly) by 

 a strong Transverse ridge upon wliich as a fulcrum the motion of the articulation is 

 accommodated ; this ridge is usually undifferentiated, but in one family it bears at 

 either enil small triangular excavations known as Supplementary ligament fossae; just 

 within the centi'r of this transverse ridge is a deep ])it, endmg bluidly, known as 

 the ligament pit; just ventral (distal) to the center of the transverse ridge is a 

 canal whicii ]>asses directly into the radial; this canal lodges the axial nerve cord 

 of the dorsal nervous system, and is called the Central canal; it is sometimes, but 

 not always, surrotmded by a raised rim; lying on either side of the central canal 

 are two shallow, usually triangular, but sometimes trapezoidal or even nearly 

 oblong or square, depressions with their bases lymg along the transverse ridge 

 and their apices directed inward, the Interarticular ligament fossse; beyond these 

 are the deeper fossie, broadly roimded to narrowly Imear, which accommodate the 

 muscles and are therefore called Muscular fossae; these are separated in tlie mid- 

 line either by a narrow ridge, the Intermuscular ridge, or by a groove, the Inter- 

 muscular groove; an<l their inner distal corners are rounded off so as to form a 

 more or less deej) Intermuscular notch. 



Within the radial pentagon, or the circlet formed by the radials in situ, there 

 is, in the oligophreate comatulids, a more or less dense secondary deposit of cal- 

 careous matter forming what is known as the Central plug (see fig. 11, p. 65). 



The centrodorsal is more or less excavated internally so as to accommodate the 

 chambered organ and accessoiy structures; the size of this cavity is variable; it is 

 verj' large in the macrophreatc species, so that in some cases the centrodorsal is 

 reduced to a mere shell, but it is small in the oligophreate species (see figs. 

 267-273, p. 259, oligophreate species, 286-291, p. 262, macrophreate species). 



IDENTIFICATION OF RECENT COMATULIDS. 



While the keys given for the determination of the genera and species of coma- 

 tulids are ample for rapid and correct identification, as is the case with other groups 

 a certain amount of familiarity with the animals is essential in order that the differ- 

 ential characters given in the keys may be appreciated in theu- true relative value; 

 much confusion may, however, be avoided if certain lines of procedure be followed 

 which, though as nearly as possible followed in the keys, are worthy of special 

 emphasis. 



Tlie first structures to be examined in the determination of an unknown coma- 

 tulid are the arms ; if these do not divide at all, and the cirri are irregularly arranged 

 on a discoidal or low hemispherical centrodorsal, the specimen belongs either to the 

 Pentametrocrinidffi (5 or 10 arms) (figs. 113, 114, p. 181, 11.5-118, p. 183, 119, 

 p. 185, 120, p. 187, 121, p. 189, and 122, p. 191) or to the Zygometridse (5 arms) 

 (figs. S3, p. 136, and 184, p. 235); if the cirri are in 10 columns on a long conical 



