64 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



of pinnules, which are developed as genital pinnules. Similarly two representatives 

 of the family Charitometridae, Poecilometra acoela and Pachylometra angusticalyx, 

 sometimes lack the ambulacral groove on as many as 20 pairs of proximal pinnules, 

 and Gislen found that in Diodontometra the first 9 pairs of pinnules, which have become 

 transformed into broad joints protecting the gonads, have no ambulacral groove. 



In all these cases the reduction of the ambulacral groove is the same on all the 

 arms. The comasterids, which show far reaching reduction in the apparatus for 

 obtaining food, have gone in another direction. . Gislen said that these animals are in 

 a state of transition from a radial type they are changing to a pronouncedly bilateral 

 type. The mouth moves forward in the anterior radius, or toward the right anterior 

 interradius, and often takes a marginal position on the disk. The anal funnel assumes 

 the central position of the mouth, in consequence of a lengthening of the intestine. 

 The anterior arms lengthen and their pinnules become long and slender, like the 

 distal pinnules of Tropiometra. The posterior (aboral) arms are shortened and have 

 limited longitudinal growth. They become thicker, stouter, and clumsier, according 

 to Gislen perhaps, though not very likely, on account of the abundant formation of sex 

 products in them. A cooperating factor in the enlargement of the joints is certainly 

 to be found in the strengthened muscular and ligamentary connections between the 

 brachials. As a result of this division of labor between the anterior and posterior 

 arms the ambulacral furrows lose their radially symmetrical arrangement and are often 

 present on the anterior radii only. It is to be observed that the ambulacral furrow 

 really disappears, and is not found remaining "in a closed groove." 



Great variation prevails even in the same species. In a number of 10-armed 

 specimens of Comatula pectinata and its variety purpurea from Cape Jaubert, Western 

 Australia, there is no ambulacral furrow en from none to four of the posterior arms ; hi a 

 Comanthus parvicirra from Sagami Bay, Japan, out of 33 arms 13 in the posterior 

 radii had no ambulacral groove, while 4 neighboring arms had it developed distally 

 of the fifteenth or twentieth pah- of pinnules. 



Gislen remarked that the reduction of the ambulacral grooves goes so far in the 

 Comasteridae that one is tempted to ask whether the apparatus for gathering food 

 can really be sufficiently large to collect nourishment enough for these relatively stout 

 and clumsy animals. 



In order to form an opinion of the relations existing between the apparatus for 

 collecting food and the size of the animal Gislen estimated the total length of the 

 ambulacral groove, measured the volume of the animal, and then by comparing the 

 figures determined their relation to each other. 



The total length of the ambulacral grooves was obtained by measuring the 

 length of the arms, multiplying by their number, and to the figure thus obtained add- 

 ing the product of the number of pinnules per arm, the average length of the pinnules, 

 and the number of arms. The final figure ;.',ives the approximate length of the 

 ambulacral grooves. 



Among the crinoids measured by Gislfhi Rhizocrinus lofotensis, which has only 

 four or five pairs of pinnules per arm, gave the lowest total figure. The length of the 

 ambulacral grooves in a 6-armed specimen was only 0.14-0.22 meters. Among the 

 comatulids measured a specimen of Comissia ignota minuta with 10 arms 20 mm. 

 long showed the shortest ambulacral grooves, 1.30 meters. It is about the same length 



