34 REPORT ON THE PENNATXJIilDA. 



of 0-044iii. A poiut of cousiderable interest is that at this spot the 

 stem is very distinctly quadrangular in trausvei'se section, the sides 

 being even slightly concave as in Funiculi no. This quadrangular 

 shape of the stem of Pennatuhi appears to have been hitherto very 

 generally overlooked. From the point mentioned the stem extends 

 down to the bottom of the stalk, preserving its quadrangular character 

 until very close to the bottom where it becomes cylindrical. Its 

 diameter diminishes at first very gradually, but in the bottom half- 

 inch very rapidly. On reaching the bottomi of the stalk it is bent 

 back on itself, so as to form a hook, the loop of the hook being in 

 contact with the bottom of the stalk, and the upturned limb of the 

 hook, which is ^in long, being extremely slender and only very 

 imperfectly calcified. The extreme tip is bent back a second time 

 towai'ds the lower end of the stalk for a length of about iin. 



In the rachis the stem loses its quadrangular character almost 

 immediately, becoming cylindrical ; its transverse section being 

 circular or somewhat oval (Fig. 3 c). It tapers gradually in passing 

 upwards, and on reaching a point fin. from the top of the rachis 

 bends back on itself for a length of about iin., ending in an extremely 

 slender and flexible thi'ead. 



In the second specimen, which has a total length of o^ins., of which 

 the stalk foniis the lower 2ins., and which has thirty-one leaves on each 

 side of the rachis. each of the larger ones consisting of twelve polypes, 

 the general relations are very similar. The stem is again distinctly 

 quadrangular at the junction of stalk and rachis, its sides being even 

 more decidedly concave than in the former specimen ; and the quad- 

 rangular character is preserved until very near the bottom of the 

 stalk. As before, the stem diminishes in diameter very slowly at first, 

 but rapidly in the last half inch. It extends to the bottom of the 

 stalk, and then tiirns back on itself for a length of ^in., forming a hook 

 and ending in a very slender thread. 



In the rachis the stem becomes cylindrical almost at once, and, 

 unlike the former specimen, slightly increases at first in size, its 

 greatest diameter, O-OiTin., being attained about iin. above the com- 

 mencement of the rachis ; from which point it tapers gradually to its 

 upper end. It extends up as far as the level of the eighth pair of 

 leaves, gin. from the top, and then bends back on itself, forming a 

 loop about ^in. long, and ending as before in an exceedingly slender 

 thread. 



In the Oban specimens we have been able to confirm the above 

 description to a certain extent. Owing to the thinness of the wall of 

 the lower end of the stalk, it is easy to demonstrate that the stem 

 extends quite down to the bottom and then turns back on itself for a 

 certain distance ; also, that it is thin and flexible at this lower end. 

 In the rachis it is, as shown in Fig. 3 c, oval in transverse section ; 

 and concerning its extent upwards, it appeai-s, so far as can be deter- 

 mined by external manipulation, to stop about half an inch from the top. 



