PKOF. P. M. DUNCAN ON A NEW GENUS OF MADEEPOEARIA. 27 



is long and narrow, and marked by the roots of several branchlets 

 which, have been fractured from the main mass, their axial corallites 

 being seen in transverse section. The height is 140 millim. ; the 

 breadth is rather greater than the height ; and the thickness is from 

 34 to 52 millim. 



In Milne-Ed wards and Jules Haime's rendering of d'Orbigny's 

 generic definition of Septastrcea the words " multiplication by fissi- 

 parity ? " occur. The note of interrogation is not repeated in their 

 own description of the genus in the ' Histoire JST aturelle des 

 Coralliaires.' Pissiparity occurs evidently enough in the Mesozoic 

 species without a columella and with an open axial space. It is a 

 fact, however, that there is not a single instance of progressing 

 fission in any calice of the type described by the French authors, 

 which is in the British Museum. There are no calices with a figure- 

 of-eight shape, and in none are there small and new septa starting 

 from the sides of the long septum-like structure which crosses the 

 perfect and full-grown calices. It is quite a mistake to state that 

 this striking feature has to do with fissiparity. In the specimen 

 belonging to the Scarborough Philosophical Society's Museum there 

 is some crowding of the calices at the base of one of the branchlets, 

 and the intercorallite walls thare are thin ; there is no superficial 

 groove, and the appearance of the septa resembles that of some 

 recent types which are undergoing fissiparity and in which the 

 division is accompanied by very rapid growth (PI. III. fig. 3). 

 But it is nevertheless true that these appearances may be the result 

 of irregular corallite-growth under the influence of pressure from 

 crowding. 



The instances of fissiparity are exceptional ; but the process 

 existed, for there is a fair example in the fractured surface of the 

 stem of the specimen first examined (fig. 2), but not in Milne- 

 Edwards's type ; and there is a very remarkable and suggestive in- 

 stance in the smallest of the specimens in the British Museum (figs. 4 

 and 4'). A section was cut at a slight depth, parallel with the surface, 

 so that sections of corallites were made at a little distance from the 

 calices above. It was evident, on counting the calices and the 

 sectioned corallites, that the former were more numerous than the 

 others, and the reason was because one of the sectioned corallites 

 was divided by fission (fig. 4). The commencement of the process 

 can be traced, and a nipping-in is to be seen with considerable con- 

 fusion of the septa. The long lamina crosses the corallite at right 

 angles to the commencing fission, and that is not what would have 

 been Seen had it been a factor in the process. On examining the 

 free surface of the colony corresponding to the sectioned corallite, 

 two calices are noted which have an incomplete wall between them ; 

 moreover a septum will be observed passing from one calice to the 

 other (fig. 4'). These calices are separated from their neighbours 

 by well-developed walls. It appears that the growth of the 

 dividing corallite was rapid. 



The number of the calices with the long lamina (that is to say, 

 perfect calices) varies in different parts of the colony ; and if square 

 patches are separated and the number of calices of all kinds on them 



