318 ON THE MADEEPOEAEIAN SUBEAMILT LOPHOSERIN^. 



and transverse in direction. Usually tlie synapticula remain 

 separate ; but in some instances they unite laterally, alDore 

 and below, and form a dense structure which acts as a calicular 

 wail. They do not form long and more or less vertical ridges 

 on the sides of opposed septa as in Eunginse, but one synap- 

 ticulum may be long and nearly sufficiently continuous to act 

 as a wall. 



The septa in the Lophoserinse may be solid or trabecular in 

 some parts ; some bring the subfamily in relation to the Perfo- 

 rata; but the classificatory distinction is evident. 



The calices may be surrounded by a wall, or its place may be 

 occupied by distinct synapticula or by those which have become 

 fused with age and growth. 



A true dissepimental endotheca is not present in the Lophose- 

 rinse. Minute growths of it are produced by disease. Hence 

 this absence observed by Milne-Edwards and Jules Haime is signi- 

 ficant and of classificatory value. 



The family Fuugidse is really something more than a transi- 

 tional group between the Astraeidae and Perforata : it has a great 

 individuality. Although certain forms, such as Mceanclroseris 

 and Goscinarcea, ally it to the last-named group, they need not 

 be placed beyond the Lophoserine subfamily ; and they belong 

 to it. 



The new genus Flesioseris has a true wall, and its septa are 

 like those of many Astrgeidse amongst the Aporosa; its synap- 

 ticula are well developed, and its external resemblance to Como- 

 seris of the Jurassic rocks is great. There are no endothecal 

 dissepiments. So that this genus and Loplioseris stand as most 

 typical Lophoserines, and nearer the AstrseidsD than Mceandro- 

 seris and Goscinarcea. 



Milne-Edwards and Jules Haime wrote about the incomplete 

 synapticula of Siderastrcea. In 1863, in describing Siderastrcea 

 grandis of Jamaica (Quart. Journ. Greol. Soc. May 1863, p. 441, 

 pi. xvi. figs. 5 a and 5 5), I noticed the synapticula ; and, later on, 

 Pourtales gave a beautiful plate proving their true nature beyond 

 a doubt. The genus has endothecal dissepiments ; and, as in 

 the case of JEcliinojpora, this structural character removes it from 

 the Fungidse proper. 



As JferuUna turns out to be a serial Astraeid, there is room 

 for a group between the Aporosa, Astraeidse, and the Fungidse ; 

 Siderastrcea and Mchmo;pora there find a resting-place. 



