452 Mr. J. Thomson and Dr. H. A. Nicholson on the 



by minute vesicular tissue, the vesicles of which are arranged 

 in oblique rows directed outwards and upwards. 



The corallum in Glisiophyllum is invariably simple ; and 

 only those examples in which the calice is well preserved afford 

 any external indication of its singularly complex interior con- 

 stitution. The form of the corallum is usually more or less 

 conical, generally curved, and rarely attaining a large size. 

 The epitlieca is variable in thickness, and is usually marked 

 by numerous encircling striae and annulations of growth. 



The calice is circular, sometimes deep and sometimes shallow, 

 its margins thin, or at other times thick and everted. From 

 the centre of the floor of the calice rises a prominent conical 

 boss or tent-shaped mass, the summit of which shows a longi- 

 tudinal crest, while its surface is marked, in well preserved ex- 

 amples, by spirally twisting or straight ridges which pass from 

 its base to its crown (PI. XXII. %. 1). This tent-shaped boss 

 has been regarded as the upper extremity of a gigantic colu- 

 mella or pseudo-columella ; but its true constitution is rendered 

 apparent by means of transverse and longitudinal sections. 

 When examined in longitudinal section, the centre of the boss 

 is seen to be formed by a columellarian lamina, which extends 

 as a continuous line (PI. XXI. figs. 2 A & 4) from the bottom of 

 the visceral chamber to the summit of the calicine dome, where 

 it appears as the median crest above spoken of. This median 

 crest has been stated to become directly continuous with one 

 of the primary septa ; but we have been unable to confirm 

 this observation ; and, in point of fact, sections indicate clearly 

 that the columellar line is confined to the centre of the visceral 

 chamber. The outer portions of the boss and its downward 

 continuation are minutely vesicular, and are formed by the 

 intersection and combination of two different sets of plates. 

 One of these can only be seen in cross sections (PI. XXI. 

 figs. 1-3), and consists of a series of vertical, more or less 

 spirally twisted lamellae, which have usually been regarded as 

 the inward prolongations of the primary septa. They are, 

 however, in reality, wholly independent of the septa, with 

 which they are never connected, except it be by means of a 

 few flexuous and delicate dissepiments, to which they are 

 always much inferior in number. It is the plates of this series, 

 also, which form the radiating ridges on the surface of the 

 conical boss within the calice. The plates of the second 

 series are seen in transverse sections (PI. XXI. figs. 1-3) as 

 a series of delicate, somewhat curved laminae, which run 

 across and closely intersect the spaces between the lamella 

 of the first ox vertical series ; but they are most characteristi- 

 cally displayed in longitudinal sections. When viewed in this 



