Structure in the Devonian Limestone, &c. 415 
Goldfuss’s ‘ Petrefacta.’ De Blainville rightly describes their 
structure as being composed of “un tissu entitrement réti- 
culé” (Man. d’Actinol. 1834, p. 537); and D’Orbigny, “ que 
leur tissu a toujours été calcaire et pierreux’’ (Cours de Géol. 
1849, vol. i. p. 208): two higher authorities it would be diffi- 
cult to find. 
In both these characters I agree; but we must go a little 
further and endeavour to find out what the nature of the orga- 
nism was that really built them. 
We may notice, then, that in all the structure is laminar, 
and the base a convoluted fibre like that of Parkeria, which, 
when magnified, would, on the unbroken surface, represent 
the convolutions of a brain. This was the ccenenchymal 
skeleton, while in the midst of its labyrinthic structure was 
an equally tortuous canal-work which held the coenenchyma 
itself, and which, again, communicated with the exterior 
through the sulcate lines between the convolutions. Thus 
the sulcate lines in their turn became part of the labyrinthic 
tubulation as the organism added layers to the surface of its 
corallum, just as in Millepora alcicornis. My. Chas. Moore’s 
specimen, to which I have alluded (‘ Annals,’ 1878, vol. i. 
p- 310), shows this most satisfactorily ; for, like an uninfiltrated 
Parkeria, the tubulation is all empty. 
So far nos. 1-3 inclusive resembled Parkeria. But this 
structure in no. 4 was accompanied by straight tubes radiating 
from the centre at the commencement, increased in number 
with the increasing growth and extended circumference of the 
cylindrical corallum, terminating on the surface in the little 
holes or calicles above mentioned, among which the shallowest, 
of course, belong to the last layer, while the deepest belong 
to the first; these tubes are about 1-48th of an inch in 
diameter, and do not appear to have had any tabule. Much 
the same structure appears to have existed with the calicles 
in the agaricoid and vasiform varieties; only, of course, the 
distance from the surface to the centre in the latter is far less 
than that of the cylindrical form last mentioned. 
Further, we observe that the ecenenchymal skeleton presents 
a layer of finer structure of the same kind on the surface, simi- 
lar to that of Millepora alcicornis (just as the circumference 
of a tree &c. presents a finer structure on the growing surface 
than further in towards the stock) ; while just below this runs 
the ‘deep horizontal canal-system”’ of Moseley, to which I 
have alluded (‘Annals,’ 1878, vol. i. p. 305), which in some 
specimens of nos. 2 and 3, where it has been half rubbed off, 
shows the walls of the canal and its interior, now chalce- 
donized, as I have described it on the surface of Millepora 
Woodwardii (‘ Annals,’ 1877, vol. xix. p. 65), while here 
