50 PROCEEDINGS OF THE ACADEMY OP [1886. 



described in my before-mentioned essay under the name of Stro- 

 matopora vesiculosa, as their vertical sections appear as an accu- 

 mulation of layers of cystose cells, rather than as a succession of 

 continuous laminae, separated by interstitial spaces. The demar- 

 kating walls of the cells appear in vertical sections as the meshes 

 of a net-work bounded by lines diagonally intersecting the direc- 

 tion which the laminae should have if they formed continuous 

 leaves. In some of these specimens a banded structure is ol)serv- 

 able, whereby from five to eight layers of vesicules are grouped- 

 together to form a broader band, which then is sharply defined by 

 a straight continuous division line from the adjoining similar 

 bands above and below it. 



Other specimens of this vesiculose species of Stromatopora 

 are found intergrown with stems of Syringopora /ilifo7^mis, 

 sometimes also with stems of Diphyphyllum multicaule, exactly 

 in the same manner as in the previousl}' mentioned Devonian 

 forms, which have been subordinated by the authors to Gauno- 

 pora, supposing the enveloped stems to be part of the skeleton 

 of Stromatopo7'a. 



Among the Devonian representatives of his genus Glathro- 

 dictyon, Mr. Nicholson describes, under the name G. cellulosum, 

 a form which I had in my paper named Stromal, cellulifera. It 

 occurs in the corniferous limestone of Port Colborne and else- 

 where, in association with another form which I called Str. 

 invoxjinata, but which, it seems, did not come under his observa- 

 tion. C. cellulosum is described as growing in irregular expan- 

 sions of considerable size, formed of a succession of horizontal 

 lamelhie, about four occupying the space of one line, and so 

 inflected as to form complete or incomplete partitions, which 

 divide the interlaminar spaces into a number of irregularly oval 

 vesicules, about three occup3ang one line. Surface tuberculated. 

 or granulated ; the tubercules occasionally perforated. In the 

 specimens found in the same locality the surface, instead of 

 being tuberculated or granulated, as above described, is covered 

 with roof-like, elongated, laterally uniting crests, forming in this 

 way a network, enclosing shallow subpolygonal cell-pits, some- 

 what similar to the shallow pits of a minute compound star-coral, 

 as the margins of these crests on weathered silicified specimens, 

 being of porous texture, appear sometimes crenulated, which 

 crenulation increases the resemblance. The other form allied to 



