138 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



on the next pinnule on the opposite side of the arm, the pinnule corresponding to 

 P 2 in other forms. Thus the arrangement of the pinnule pairs in the genus Eudio- 

 crinuft is just the reverse of what it is in the other comatulids, P, being unique in 

 its structure when compared with the other pinnules succeeding, and the pinnule 

 pairs consisting of P a and P,, P& and P s , P c and P,, etc., instead of consisting of 

 P, and P n . P 2 and P&, P 3 and P r , P 4 and P,/, etc. But if we disregard the arm struc- 

 ture of Eudiocrinus and consider only the pinnules, we see an arrangement identical 

 with that on the free undivided arms of other types, for P c agrees with P t , P with 

 P 2 , P,, with P 3 , and so on. Thus the pinnulation in Eudiocrinus is that of a normal 

 undivided arm and is quite independent of the unique arm structure. 



In many of the multibrachiate comatulids, well illustrated in Pontiometra 

 anderson! (figs. 275-277 p. 213) and in Lamprometra protectits, the proximal 

 pinnules on the outer side of the outer arms arising from each radial are much 

 more developed than the pinnules on the inner arms and on the inner side of 

 the same arms, though their general characteristics are the same. In 20-armed 

 species only the lower pinnules on the outer side of the outer arms are enlarged, 

 but in 40-armed types a similar, though not quite so marked, enlargement of the 

 pinnules on the outer side of the innermost derivatives from each of the IIBr 

 series also occurs, and in species with a great number of arms the pinnules on 

 the outer side of the outer arms of each group of four are enlarged, this enlarge- 

 ment diminishing in degree toward the inner part of the ray according to the 

 scheme given on page 124. 



In the ontogeny of the comatulids pinnules first appear at the tip of the 

 growing arm after the arm has reached a considerable length (figs. 1204, pi. 33; 

 1207, 1210, pi. 34: 1217-1219, pi. 35). P t is not formed until after from two to 

 half a dozen or more of the outer pinnules have reached a considerable size, and 

 the pinnules of the fourth and four or five following brachials not until after 

 P! is relatively well developed (fig. 1220. pi. 36). 



Deficient pinnulation in the proximal portion of the arms is characteristic 

 of many species. In some it is constant and invariable, in others more or less 

 inconstant, while in still others it only occurs sporadically. 



In the Comasteridse the species of the genus Comatilia (figs. 183, p. 98, 

 and 2G4. p. 207) lack the second, third, and fourth pairs of pinnules, though the 

 pinnules of the first pair, P, and P , are present and highly developed. This 

 condition is quite comparable to that of the young after the development of P , 

 and may have arisen through a sudden arrest in pinnule formation. 



In the genus Atelecrinus there are no pinnules on the first 10 or 12 brachials. 

 In this genus the arms are very closely appressed, and the formation of pinnules 

 on the lower brachials is thus prevented. In the allied Atopocrinus (part 1, 

 fig. 227. p. 245), in which there are only five arms, all the pinnules are present. 



In T haumatocrinus (part 1, figs. 115, 116, 118, p. 183) the first pinnule is 

 always on the second postradial ossicle, but in nearly all the species of the closely 

 related Pentametrocrinus (part 1, figs. 120, p. 187, 121, p. 189; exception, fig. 

 119, p. 185) the lowest pinnule is on the fifth brachial; that is, on the epizygal of 

 the first syzygial pair, the normal first pinnule being absent. 



