1916] The Ottawa Naturalist 71 



of crinoids were imagined to have been uniserial. However, in 

 times preceding the advent of the actually known paleozoic 

 crinoids, adjacent uniserial arms were supposed to have united 

 laterally in pairs in such a manner as to give rise, first, to bi- 

 serial arms, and, later, to pseudo-uni serial ones. According 

 to this theory, the pinnules of the theoretical uniserial arms 

 might be arranged in a single series along one side of the arm, 

 while the pinnules of the pseudo-uniserial arms should occur 

 in two series, successive pinnules being attached alternately 

 to opposite sides of the series of arm ossicles. If the food- 

 groove along the ventral surface of the crinoid arms be regarded 

 as originating along the line of junction of the two imaginary 

 primitive uniserial arms, this food-groove might be retained 

 in pseudo-uniserial arms originating from biserial forms, but 

 need not be present in the imaginary primitive uniserial arms. 



The views favored by Clark, and the various possible de- 

 ductions from them, are interesting. They would be more 

 interesting if they found support in the probable phylogeny 

 of fossil species. It must be conceded, however, that' in the 

 earliest known representatives of the crinoids, the primary 

 radials and primibrachs of Clark already were united laterally 

 so as to present an initial series of five, instead of ten arms, as 

 demanded by Clark's theory, and all the arms bear food-grooves. 

 Moreover, even the earliest known biserial arms are more or 

 less uniserial at the base. 



2. Uniserial arms and pinnules in Comarocystites. 



In the absence of anything corresponding to the supposed 

 primitive arm structure of crinoids, among known Crinoidea, 

 it may be interesting to note that, among the Cystidea, the free 

 arms of Comarocystites are uniserial (Plate III), do not bear a 

 food-groove along the ventral side, and support pinnules ar- 

 ranged in a single row along the right side of the arm (the ven- 

 tral surface being directed away from the observer, and the 

 distal end of the arm being directed upward) ; moreover, the 

 pinntiles consist of a uniserial row of ossicles. In a similar 

 manner the uniserial row of plates supporting the recumbent 

 food-grooves of Amygdalocystites (Canadian Organic Remains, 

 III, 1858, plate VI), also might be regarded as uniserial arms, 

 bearing a single row of uniserial pinnules along the right side 

 of each arm. It is probable that Canadocystis (Bulletin 80, 

 N. Y. State Museum, 1905, pp. 2 73, 2 74), had an arm structure 

 similar to that of Amygdalocystites. It must be admitted, how- 

 ever, that these forms are not normal cystids. The possession 

 of uniserial pinnules in Comarocystites and Amygdalocystites is 

 sufficient to indicate this. Canadocystis probably also had uni- 

 serial pinnules. However, none of these genera could have 



