2 TH. MORTENSEN, (Schwed. Südpolar-Exp. 



is in conformity with the well known fact that viviparous forms are specially numerous 

 in the Antarctic Regions. That each of the three forms has a separate way of caring 

 for the brood adds, of course, considerably to the interest attached to them. 



I beg to tender my sincere thanks to Dr. K. A. Andersson and Professor O. 

 NORDENSKJOLD for leaving me this exceptionally interesting material. 



The five species contained in this collection are: 



1. Notocrinus virilis MRT.SN. 



2. Isometra vivipara Mrtsn. 



3. Thauniatometra nutrix n. sp. 



4. Anthometra adriani (Bell). 



5. Promachocrinus kerguelensis P. H. CARPENTER. 



Fam. Notocrinidae MrtsN. 

 I. Notocrinus virilis Mrtsn. 



PL L Figs. 1—5; PL IL Figs. 1—4: PL III— IV. 



Notocrinus virilis. Th. Mortensen. Notocrinus virilis n. g., n. sp., a new viviparous Cri- 

 noid from the Antarctic Sea. Preliminary Notice. Vid. Medd. Dansk 

 Naturh. Foren. Bd. 68. 1917, p. 206 — 7. 



Centrodorsal rather large, conical; the cirrus sockets arranged in ten columns of 

 three or four each, separated by a naked space in the radii. (PI. II. Fig. i). The 

 dorsal pole is rough. In the younger of two specimens, a half grown and a full grown 

 one, which were sacrificed for a more detailed study, a fine pore is observed in the 

 middle of the apex, surrounded by five peripheral, radially placed pores; only the 

 central pore perforates the wall of the centrodorsal completely, the peripheral ones 

 having apparently obliterated their connection with the central cavity. In the larger 

 specimen the peripheral pores have completely disappeared, while there is still a distinct 

 trace of the central pore. — The presence of these pores (for the central canal of the 

 stalk) in half grown specimens would seem to indicate that the stalk is retained for 

 a considerably longer time than usual in Comatulids. 



The upper edge of the centrodorsal is, in the larger specimens, rather deeply in- 

 cised below the radiais, the interradial corners being quite prominent. The ventral 

 side of the centrodorsal has a large pit in each radius (PI. II. figs. 2 — 3); it is very 

 deep, as deep as the central cavity, thus separating the two series of cirrus-canals of 

 each radius. There is a very distinct, Icaflike basal groove (PI. 11. Fig. 3). The 

 ventral median edge of the centrodorsal does not continue inwardly so as to form a 

 diaphragm. 



