298 DESCRIPTION OF CHALK CORALS. 
In the mode of growth, and in the existence of lateral bands without visceral 
cavities, this coral resembles Ptilodictyum lanceolatum'; but that fossil is tubular, 
and is distinguished by its subordinate structures from the subject of the present 
inquiry. For certain cretaceous and subcretaceous Bryozoa, Herr Roemer has 
proposed the genus Rosacilla; and one chalk species, Ros. serpuleformis’, is repre- 
sented with a marginal lamina resembling that of the zoophyte under consider- 
ation, also with an approximation towards a transverse arrangement of apertures ; 
and the fossil is stated to be formed of fine, long tubes, invisible externally: the 
genus, moreover, is placed among the Tubuliporide, and described as consisting of 
circular, less frequently of irregular incrustations, composed of one or more layers 
of cells (tubes) with projecting mouths. These structures afford some points of 
agreement with those of the chalk coral, and the concealment of the tubes ex- 
ternally does not appear to have resulted from a thickening due to age—the mode 
of growth also, as well as the occasional existence of more than one layer of cells or 
tubes, cannot likewise be regarded as satisfactory generic distinctions. Of the 
five cretaceous species mentioned by Herr Roemer, four however have the form 
of the tube visible externally, three of the number having besides a distinct 
furrow between the cells. This character indicates, it is conceived, the want 
of an external lamina similar to that of Mr. Dixon’s fossil, and consequently 
that an essential difference existed in the composition of the polypes; but 
it must be stated that in two species of Rosacilla the tubes are described as 
streaked across or obliquely (‘‘ queergestreift ”’), though without any allusion to 
capillary tubes constituting a structure independent of the integument of the 
visceral cavity. The mode of producing additional tubes is not mentioned ; nor 
the extent to which those composing one layer overlie each other ; but the en- 
larged figure of Ros. serpuleformis exhibits possibly in the transverse fracture a 
considerable overlapping, no allusion being made in the description to successive 
layers. In the chalk fossil the cavities apparently covered each other very 
slightly, if at all, the additions, so far as could be ascertained (fig. 2 a), springing 
either from the point where the preceding tube inclined outwards, or from the 
portion of the side immediately adjacent to it. The latter character will most 
probably be considered of little importance, the composition of the genus not 
* See Goldfuss, Flustra lanceolata, p. 104, tab. 37. f. 2; also Hisinger’s Lethwa Suecica, p. 104, 
tab. 29. f.10; likewise Sir R. I. Murchison’s Silurian System, p. 676, Ptilodictya lanceolata, pl. 15. 
figs. 11 tolle; and Lieut.-Col. Portlock’s Geol. Rep. on Londonderry, &c. p. 339, pl. 21. figs. 3 & 6. 
* Verst. Norddeuts. Kreidegebirges, p. 19. tab. 5. figs. 16 a, b. 
