MADREPORARTAN FAMIIiT THE FU]S"GHD^. 139 



primary groups of Madreporaria comprising families. The Fun- 

 gidffi (with tlieir synonyms of Fungacea, Verrill, and Fimgiacea, 

 Klunzinger) have not the same zoological significance as the 

 groups just mentioned, and cannot be placed except as a family 

 divisible into subfamilies and genera. 



Since Dana established the family Fungidse, some fossil and 

 recent genera which had been classified with the Aporosa have 

 been admitted into it ; but that affords no reason for an alte- 

 ration of name. Moseley retains the recognized position of the 

 Fungidse, and so did M. de Pourtales. 



The Fvngidcs as differentiated hy MM. Milne-Edwards a7id 

 Jules Hainie *. — " One of the general and most striking characte- 

 ristics of the species which form this family is the short and 

 expanded growth of the corallum, whether it increases by gem- 

 mation and becomes a compound form, or whether it remains a 

 simple one. But this tendency to a more or less horizontal de- 

 velopment of shape is not invariable and absolute ; and by itself 

 it could not afford a satisfactory differentiation of the type, were, 

 this external character not associated with an internal struc- 

 ture of a very great importance. We have already seen that the 

 interseptal loculi are either vacant down to their bases in such 

 groups as the Turbinolidse and Dasmidse ; or are closed at certain 

 heights by lamellar dissepiments, as in the Oculinidse, and princi- 

 pally in the Astrseidse." 



"The Fungidae present a new disposition in the structure of 

 their interseptal loculi, which is not found in other groups. The 

 dissepimental tissue is completely deficient in their interseptal 

 loculi as in the Turbinolidse ; but the dermal sclerenchyma of 

 the sides of the septa becomes developed in places and extends 

 beyond the septal laminse, forming projections like warts or 

 tubercles, ivhich groio towards those of the opposite sides of the next 

 septum and fuse. It follows that the interseptal loculi and the 

 visceral cavity in that position are more or less traversed, but 

 never completely closed, by kinds of bars often of considerable 

 dimensions. This interseptal structure differs greatly from 

 the endothecal structure of the Astrseidae, both in its nature 

 and analogies. We may consider the structures forming it to be 

 analogous to extremely developed granules such as are seen in 

 most of the septa of the Turbinolidae, Astraeidse, and Madrepo- 

 ridse. These transversely placed organs, which we have proposed 

 * Hist. Nat. des Corall. vol. iii. p. 1 et seq. 



