66 CONTRIBUTIONS TO CANADIAN PALAEONTOLOGY. 



Donald, B.C. (Silurian) they do not exceed *7 mm. by '45 mm. in size ; 

 between these extreme forms may be found specimens with corallites of 

 various sizes. Corallites that are markedly quadrangular in transverse 

 section are met with, as well as those, in diffferent specimens, that are 

 almost quadrangular, oval, both broadly and narrowly oval and circular, 

 whilst both circular and oval corallites frequently occur in the same speci- 

 men. H. gracilis, Hall, was separated as a distinct species principally on 

 account of its corallites being " quadrangular " as seen in horizontal section; 

 as will be seen further on, there are other reasons for still considering it 

 as distinct from the typical form of //. catenularia, at least as a variety 

 indicative of distinct geological horizons. 



If after finding so many transitions in outward appearance between 

 extreme forms of this coral as to render a specific classification upon this 

 basis alone extremely difficult, if not impossible, one passes on to an 

 examination of its internal structure, difficulties arising from trans- 

 itional changes are again met with. Nicholson has adopted a classifi- 

 cation of the different forms based on a microscopical study of the inter- 

 nal construction, and has separated them into two groups ; in one he 

 places those forms in which the corallites are found apparently con- 

 tiguous, in the other those having a tabulate space of variable width 

 separating the corallites. In his " Manual of Palaeontology," 3rd ed. 

 vol. I., p. 339, Dr. Nicholson observes " The species of Haly sites may 

 be divided into two groups, according as the corallum is composed through- 

 out of corallites of one size or consists of two sets of corallites of different 

 sizes. The common H. escharoides of the Silurian rocks is an example of 

 the forms in which the corallites are similar. On the other hand, in the 

 familiar //. catenularia of the same formation, the corallum consists of 

 large corallites separated by the intervention of small closely tabulate 

 tubes." An examination of the structure of Canadian specimens reveals 

 the fact that in different specimens the small closely tabulate tubes or 

 tubules described by Nicholson are not constant in their proportionate 

 size to the corallites and differ in size from those which are wider than the 

 corallites themselves to those which have a scarcely appreciable width, 

 making a gradual transition to the forms in which the tubules appear to 

 be obsolete. In a specimen from l'Anse au Gascon, Baie des Chaleurs 

 (Lower Helderberg) the tubules sometimes exceed the corallites in width 

 and are furnished with very close set tabulae which have an almost vesic- 

 ular appearance on account of the introduction of secondary tabulae ; the 

 tabulae are depressed in the centre and are bent downward at their edges. 

 In the Niagara formation in Ontario examples occur in which the 

 tubules are slightly less than one half the width of the corallites, but the 

 majority of the specimens from this formation in Ontario, at Lake 



