534 MR. J. STANLEY GARDINER OX FuiJ-GiD [June 21, 



true columella, but to be formed on the top of fused trabeculae 

 from the septal edges. 



In P. haindana (PI. XLV. fig. 1) the septa are evenly covered on 

 the upper surface M'ith blunt, subequal spines, which are much 

 more numerous than in P. ohtusangula, but have given rise to the 

 same ridging of the septa, although less markedly. Intermediate 

 to the axial fossae the synapticula form a very thick vertical row 

 between the several septa, reaching right up to their edges below 

 the spines and simulating a wall. Commonly, close to this, on 

 either side, is another row of large synapticula. Deeper in the 

 colony these rows fuse, owing apparently to the thickening of the 

 corallum between and the formation of trabeculse from the sides 

 of the septa, a single, thick, slightly perforated wall resulting. 

 Synapticula are found also bet\\een the septa within this false 

 wall ; they are much thinner and less elongated than those which 

 form the wall, but placed in the same waj^ in vertical series between 

 opposing thickenings of the septa. The axial fossa is closed beloM' 

 by trabeculae from the septal edges : commonly a central spine can 

 be distinguished, surrounded by a circle of spines, corresponding 

 more or less in number with the septa. 



I have also ascribed to the genus Psammocora three new species, 

 which I propose to call P. profandaceUa, P. su])ei-Jicialis, and 

 P. savigniensis, which agree with it in the general characters of 

 their septa, the arrangement of the synapticula, and the formation 

 of a false columella by the fusion of trabeculae from the septal 

 edges. 



P. superjicialis (PI. XLY. fig. 2) resembles Mci'androsens bottce, 

 Rousseau, and differs from P. haimiana, in that its calices are 

 not nearly so completely circumscriljed, and are arranged more or 

 less in series : the septa, too, are more regularly ridged, more 

 perforate, and usually continuous between the calicular fossae. 

 There is, as in P. haimiana, between neighbouring fossae a distinct 

 central row of synapticula, which, however, do not rise so high as 

 to give from surface-view the appearance of a wall. The rows on 

 either side are well marked, but are not generally visible from 

 the surface ; deeper in the corallum, however, they are fused 

 with the central row, and form a wall precisely similar to that of 

 P. haimiana. The axial fossa is closed in below, both in Mmindro- 

 seris hotta;, as described by Duncan \ and in P. superjicialis as in 

 P. haimiana, by trabecule from the septal edges. 



P. profumlacella (PI. XLV. fig. 3) very closely resembles PJesio- 

 seris australice, Eousseau, indeed only differing from it, so far as 

 the description goes, in having its calices rather deeper, with a 

 considerably larger number of septa and less regularly arranged in 

 series. Plesioscris australice has, however, according to Duncan'', 

 a true wall. In P. profundacella, as also in a specimen named 

 Plesioseris, apparently by Duncan, in the British Museum, there is 

 no true wall, but between the fossae a row of synapticula, very 



^ Journ. Linn. Soc, Zool. xvii. p. 308. 

 ^ Loc. cii. p. 309. 



