232 TABULATE CORALS. 



the following remarks : " The tubes in the different species 

 vary from one quarter of a line to nearly one line in breadth ; 

 they are very long, and are most frequently united throughout 

 laterally, forming massive coralla, resembling more or less those 

 of Favosites and ChcEtetes ; sometimes, however, they are united 

 in single intersecting series, as in Halysites catenulata, Linn. ; 

 not unfrequently, too, the tubes are isolated, or only united at 

 irreo^ular intervals, thus forming- loose fasciculated coralla, re- 

 sembling certain forms of Syringopora." 



Professor Safford further states that the isolated tubes are 

 nearly quadrangular, with more or less rounded angles, and 

 with a slight external longitudinal depression opposite to each 

 of the four septa ; the walls are more or less rugose ; and in- 

 crease is by fission of the old tubes. Only one specimen was 

 seen in which tabulae could be detected ; and in this they were 

 confined to one end of the mass, and were distant from one 

 another about twice the width of the tubes. 



The genus Tetradium is regarded by Safford as intermediate 

 between the Favositidce and the Rtigosa, the quadripartite char- 

 acter of the corallites placing it in the latter group. 



Taking such a well-known species as the 7". miwits, Saff., of 

 the Cincinnati Group, as the type of the genus Tetradium, we 

 find that the corallum (fig. 33, a) is massive, hemispherical, 

 or irregular in shape, and composed of closely amalgamated, 

 slender, prismatic corallites, which diverge from the base or 

 from an imaginary axis, and are not arranged in superimposed 

 layers.. No general epitheca seems to be present. The coral- 

 lites are in close contact throughout their length, and their walls 

 appear to be entirely imperforate. Some doubt, however, must 

 in the meanwhile remain upon this point, owing to the fact that 

 all the microscopic sections I have made (of specimens derived 

 partly from Ohio and partly from Canada) show the walls to have 

 undergone a very peculiar change. The walls of the corallites, 

 namely — though the specimens appear otherwise to be quite 

 unaltered — have the normal granular carbonate of lime which 

 composes the Ccelenterate skeleton replaced by crystalline cal- 



