306 Mr. H. J. Carter on the Stromatoporldge. 



in this state they must not be confounded with that common 

 calcareous white mycelium which, either Foraminiferous or 

 Saprolegnious, pervades almost every marine calcareous 

 structure. 



That, however, the vessels of the " proliferous membrane " 

 are sometimes calcified is worth remembering, as it may 

 hereafter explain how, in the fossil species, they appear some- 

 times as mere spaces, whereby they may often be easily 

 overlooked, and at others in the form of calcijied canals. 



The vascularity of the " proliferous membrane," which has 

 been examined by Mr. Moseley in the recent state, and iden- 

 tified with the " hydrophyton " of Allman (coenosarc auct.), 

 is beautifully represented in his figures 12 and 16, plate iii. 

 {oj}. et loc. cit.), where the vessels are shown to be surrounded 

 and filled respectively with ectodermic and endodermic cells of 

 different kinds which produce the various elements of which 

 the Millepore is composed. 



Having thus pointed out the source from which the coral- 

 lum of MiUepora alcicornis is derived, I have now briefly to 

 allude to the composition of its ultimate structure; and for this 

 purpose let us assume that a branch has been broken off" from 

 the main specimen, and that we are examining the vertical 

 fracture (for this is preferable to any other method, as involving 

 the least destruction of the more delicate parts). We may 

 observe that the branch is marked by an axial, a middle, and 

 a superficial structure, all three of which are differentiated by 

 the following peculiarities. The " axial structure " is here 

 represented by a number of minute holes in juxtaposition 

 (that is, a cribriform coenenchyma) , in which the surface of the 

 hard parts towards the holes or spaces is more compact than 

 the rest ; that is, the surface is more opaque and whiter 

 than the interior, which, on the other hand, is composed of 

 more transparent calcite. This, again, which will be generg,lly 

 found to be the case with the coenenchyma, is worth remem- 

 bering, because it will lead to the explanation of a similar 

 appearance in the elements of the fossil structure, which other- 

 wise might be set down as originally hollow or filled with 

 coenosarc. The " middle structure," on the other hand, is 

 more compact, and its coenenchyma composed of a curvilinear 

 element or fibre, moulded over a vermiculo-reticulate coeno- 

 sarc, whose tortuous anastomosing canals alone remain in the 

 dried corallum, more or less obliterated here by thickening. 

 This basic structure, again, is traversed by tubular spaces at 

 variable distances from each other, which radiate from the 

 " axial structure " towards the circumference of the branch 

 (' Annals,' 1878, vol. i. pi. xvii, fig, 5), where they end in the 



