44 SMITHSONIAN INSTITUTION 



In a few species (mostly in the genera Stephanometra, Colobometra, and Oxymetra) 

 the pinnules of the first two or three pairs are stout, extremely stiff and spine-like, sharp 

 pointed, composed of very few elongated segments, and project diagonally outward over 

 the disk which is thus protected by a mass of sharp thorns. 1 



Usually, however, the protective function is confined to the pinnules of the second 

 and one or two following pairs, the pinnules of the first pair remaining many-jointed, 

 delicate and flagellate, tactile organs. 



In most of the genera of the Thlassometridce (as in Thalassometra) and in certain 

 species of Himerometra, the pinnules of the first pair are very stout for most of their 

 length, but end in a delicate flagellate tip, thus combining both tactile and protective func- 

 tions in the same organ. 



In the Comatulids whenever the division series consist of four segments, the second 

 always bears a pinnule; whenever they consist of three segments (the palmar and sub- 

 sequent division series in Capillaster and Nemaster), the first bears a pinnule. The 

 pinnules borne on the division series always are of the same character as the first pinnule 

 of the undivided arm, but with its distinctive features accentuated ; thus if the first arm 

 pinnule is larger and stouter than the second, the palmar pinnules will be in turn slightly 

 larger and stouter than the first pinnule, and the distichal pinnules slightly larger and 

 stouter than the palmar; if the first arm pinnule is smaller and weaker than the second, 

 the palmar pinnules will be still smaller and weaker, and the distichal pinnules smaller 

 and weaker than the palmar. 



As the surface of the disk ordinarily reaches the level of the second brachial, it 

 naturally follows that before becoming free the pinnules of the division series must run 

 along over the dorsal perisome for some distance; but owing to the deep interradial 

 incision of the disk in most multibrachiate forms this means little more than that these 

 pinnules run diagonally upward to and beyond the edge of a much thickened brachial 

 perisome. 



In a few multibrachiate types (as in Comaster belli and Comanthiiia schlegelii) the 

 disk is approximately circular, with the bases of the distichal and palmar pinnules incor- 

 porated in the dorsal perisome. This dorsal perisome may become heavily plated, in which 

 event the pinnule bases become invariably fixed in a pavement of very irregular polygonal 

 plates from which, however, the pinnulars are always readily distinguishable because of 

 their height, strong convexity, and regularity in size and shape ; in other words, in recent 

 types the fundamental character of the pinnule as a free organ is never in the slightest 

 degree masked by incorporation in the perisomic wall. 



It will be observed that the condition described in the last paragraph of 

 Mr. Clark's note is substantially the same as that of the fixed pinnules in the 

 various fossil genera mentioned above, except for the greater tendency of the 

 pinnules in fossil forms to merge in the perisome. 



From the various examples thus given of the form and function of the 

 proximal pinnules, it is seen that these organs frequently undergo a course of 

 development wholly different from those succeeding them at regular intervals 

 along the arms; and it is to this tendency to unequal growth that we must 

 look for an explanation of the extraordinary condition found in Gilbert socrinus. 

 Most authors who have mentioned them have treated the tubular extensions 

 in this genus as analogous to the interbrachial pores in Batocrinus and other 



1 See Hartlaub (Comatuliden des Indischen Archipels), Nova Acta Ksl. Leop.-Carol. Deutsch. 

 Akad. Naturforscher, Band 58, Taf. 3, fig. 29, 1891. 



