270 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



lumen, the angles both of the calcareous portion of the column and of the lumen 

 being directed interrudially. 



Such a column presents five radii of maximum density (directed interradially) 

 and five radii of minimum density (directed radially). Each columnal attains a 

 very strongly stellate shape wth a very great difTerence between these two series of 

 radii before the cirri begin to develop. It is thus onl}' natural that the cirri should 

 pierce the column by the path of least resistance and should reach the exterior by 

 the shortest route, emerging radially instead of interradially. 



Apparently this distortion of the column became permanently fixed in the crinoid 

 phylogeny before the inception of the degeneration of the infrabasals which we see 

 carried to an extreme in the pentacrinitcs, and especially in the comatulids, so that 

 in these groups it remained in its secondary condition without reverting to the 

 original form. 



We have already seen (p. 142) that the sjaumetry of the dorsal skeletal sj'-stem 

 and of the dorsal nerves does not correspond with that of the ventral radial struc- 

 tures, for the mid-somatic dorsal structures are interradial and the mid-somatic 

 ventral structures are radial, the two sets having swung aj)art so that their respec- 

 tive mid-somatic axes differ in direction by 36°; in other words, a torsion has 

 been introduced into the ontogeny so that in the adults mid-somatic ventral struc- 

 tures ho directly above the intersomatic dorsal divisions. Remembering this it 

 does not occasion any surprise to find in the so-called dicyclic species (for example 

 in the pontacrinites and in the comatuhds) a second torsion so that the cirri and 

 the originally mid-somatic structures of the column, instead of maintaining the 

 same orientation as regards the calyx as they do in the monocyclic forms, have 

 become shifted through an arc of 36° and have thus come to he directly beneath 

 the niidsoraatic axes of the ventral portion of the animal. Many of the hydroids, 

 alcyonariaiis and bryozoans which have adopted a plant-like habit of growth have 

 corrolatively also adopted to a greater or lesser extent a spiral arrangement of 

 their zooids upon the central rachis which is strictly comparable to the spiral 

 arrangement of loaves upon the stem of a plant, for the economic factors governing 

 the arrangement of leaves are quite parallel to those determining the arrangement 

 of the zooids. The s])iral swing through an arc of 72° assumed by the dicyclic 

 crinoids, in two stej)s of 36° each, is the logical result of the possibihty of plant- 

 like accommodation by these plant-like organisms to meet any exigency, internal 

 or external, which may arise in the course of their phylogenotic develo])nient. 



Cirri only occur in the crinoids in which group, like the central or suranal plate 

 among the echinoids, they are by no means of universal occurrence, but are found 

 only in the more specialized, and mostly in the later, tj'pes; even in groups in 

 which they are normaUy present they may be abruptly suppressed, as in the 

 Innatantes and in the adults of many comasterids. 



Their occurrence or non-occurrence usually is of great systematic interest, but 

 too much weight altogether has been placed upon it; we have seen how Ln a number, 

 of the Comasteridse they may be only developed Ln the young and entirely suppressed 

 later; in other generathey do not appear at all until vcrv late in life, as in Proisocrinus 

 (fig. 128, p. 199). 



