parasitic ivithin recent Madreporaria, 



247 



to the outside of the corals. They are also to be distinguished in fossils, 

 crowding the cavities on the outside of the coral and some of the dissepi- 

 mental spaces ; but I have failed to see them in corresponding numbers 

 and positions in recent forms. 



The oospores are large, and reach y^Vir diameter. They are 



usually globular, but sometimes slightly lobular ; and in the latter case 

 there is an evident internal grouping of granules which makes the mass 

 look like a tetraspore. They are dark in colour, and there is an external 

 cell-wall. Becoming fixed to the coral-wall, they sometimes flatten out 

 and remain as circular black spots, a long filament being continuous with 

 them and passing into the hard structure ; but as yet I have not seen a 

 satisfactory tubular prolongation of a globular spore, except in the Silu- 

 rian Cromoj)Jii/Uum, where they present beautiful tubes, 



YI. T7ie large confervoid-looJcing Filaments ivithin the Organic Basis. 



The septa of all Corals are in two longitudinal parts, which are sepa- 

 rated more or less perfectly in the median line. Hence transverse sec- 

 tions show in the septa of recent corals two cut'surfaces, each representing 

 the top of a plate or lamina separated by a linear space or by a corre- 

 sponding orgamc tissue. In some species this intermediate structure 

 reaches the siu-face at either the calicular or the costal edge of the septiun. 

 It is the relic of an involution of the dermal structures in and around 

 which the sclerenchyma was deposited. Moreover the organic structure 

 thus intermediate really passes into the septal laminae in all directions, 

 as can be proved by decalcifying them. It would appear that the younger 

 the coral is, the more distinct are these intermediate structures, and that 

 with age they become more or less filled up with carbonate of lime, and 

 may finally thus become obliterated. Certain genera of rapidly growing 

 forms appear to have the spaces more distinctl}^ developed than others. 

 . These organic structures, coming to the surface of the coral, of course 

 are, imder favourable circumstances, soon attacked by penetrating para- 

 sites j and the possibility of entry and of growth is so great in some instances, 

 that wide layers of very confervoid-looking growths flourish, and present 

 a very peculiar appearance under the microscope (Plate 7. figs. 33-35). 

 The filaments forming one or more layers are nearly parallel, are large 

 (from j-^o~o to -2oW i^ch) , rarely branched, and exhibit transverse mark- 

 ings, which are not true articulations, although the filament frequently 

 breaks off at them. The filaments are tubular, and have a distinct cell- 

 wall, which is swollen out here and there ; and their cytioplasm is aggre- 

 gated more or less regularly, and in some places resembles conidia. Little 

 processes project from the wall of the filament here and there, and 

 are of about half the calibre of the parent tube. In some filaments the 

 cytioplasm is well separated from the wall, but usually it fills the whole 

 tube. 



The jointed or articulate appearance is particularly visible when the fila- 



