chief Generic Types of Palceozoic Corals. 291 



and is composed of amalgamated and polygonal covallites ; 

 in some cases (as, for example, in L. canadense, Castelnau) 

 the corallum is partly fasciculate and partly astra^iform, or 

 different specimens may be wholly the one or the other. In 

 any case, the general form of tlie corallum does not appear 

 to be a sufficient ground for generic distinction, though the 

 fasciculate forms have repeatedly been placed in one genus 

 and the astraiform in another. 



The increase of the corallum is effected in the typical species 

 of Lithostrotion by calicular gemmation, involving only a 

 portion of the original calice, and allowing the old corallite to 

 go on growing side by side with the new one. Lateral budding- 

 is also not uncommon, the new corallites produced in this way 

 assuming a direction parallel to that of their parents, and grow- 

 ing up side by side with them. The genus Stylaxis was founded 

 by Prof. M'Coy for species oi Lithostrotion which were sup- 

 posed to increase by fission of the old tubes. The appearance 

 of lissiparous development in these cases seems, however, to be 

 really due to the fact that the young corallites produced by 

 calicular gemmation do not hinder the continued growth of the 

 old corallites, but grow up alongside of them, the two quickly 

 becoming united by their walls ; and even if fission were proved 

 to occur occasionally, it would hardly afford of itself sufficient 

 ground for generic distinction. 



The epitheca is well developed and complete, marked with 

 circular strias and shallow annulations of growth. Sometimes 

 the corallites of the fasciculate species inosculate with one 

 another. If L. Stokesi, E. & H., be rightly referred to this 

 genus, the corallites are in this species united by outward ex- 

 pansions of the epitheca; whilst in L. harmoditesj E. & H., 

 they are united together by horizontal connecting processes, as 

 in Sp-iiiffopora. There are, however, some doubts as to the 

 true affinities of these forms. 



The septa extend from the inner surface of the wall to vary- 

 ing distances from the columella. A few of the primary septa 

 sometimes actually reach the columella ; but most of them, to- 

 gether with all the secondary septa, fall short of it. The sides 

 of the septa are plain and not denticulate ; and in the majoritv 

 of cases the primary septa extend so far inwards as not to 

 leave any conspicuous space in the centre of the visceral 

 chamber occupied solely by the tabular. An inconspicuous 

 septal fossula is sometimes present, as inZ. affine, Flem. ; and 

 the somewhat aberrant L. canadense, Castelnau, is said to 

 possess several small fossettes. 



The columella is always present in each corallite, and has 

 the form of a flattened, laterally compressed, compact rod,Mhich 



