O'CONNELL, REVISION OF THE GENUS ZAPHRENTIS 185 



the genotype, and Z. centralis, E. and H. and Z. dalii, E. and H. as other 

 examples. These species differ from normal Zaphrentis and Hetero- 

 phrentis in having the septal arrangement arrested in the primitive quad- 

 ripartite manner characteristic of the young of rugose corals generally. 

 Thus two lateral or alar pseudo-fossulae are retained. There are no den- 

 ticulations or carinas on the thickened septa. The genera are Devonian 

 and Mississippian in age. 



A third generic term proposed by Simpson is Scenophyllum, with 

 Zaphrentis conigera Eominger as the genotype. This, however, has no 

 close relationship to other zaphrentoids. Finally, he proposed the genus 

 Homalophyllum for such species as Zaphrentis ungula Eominger and 

 Zaphrentis herzeri Hall, which are flattened on the side of greatest curva- 

 ture. It is not at present certain that the two species named are con- 

 generic, in spite of this flattening on one side. 



There has been little discussion of these genera during recent years, 

 until in 1908 Carruthers published a paper entitled "A Eevision of Some 

 Carboniferous Corals," in which he especially considers the standing of 

 the genus' Caninia, He went to the sources in the literature, and after 

 carefully examining not only many specimens of Caninia cornucopice and 

 allied forms, but also several hundred from the type locality, Tournay, 

 and numerous examples from the Bristol area, he gives a re-definition of 

 the genus as follows : 



"Corallum simple, turbinate and conical, often slender and cylindrical for a 

 great part of its length, 



"Major septa well developed and meeting in the centre in the lower, conical 

 part of the coral, but in the cylindrical portions usually becoming amplexoid 

 in character. 



"Minor septa of various lengths in different species. 



"Cardinal fossula variable in extent, characteristically limited by tabulse 

 only, at the inner end, and with flanking septa loose or disconnected. 



"Tahulce well developed, but variable in regularity ; they may be highly 

 arched and vesicular. A marginal ring of more or less vertical dissepiments, 

 usually thin and delicate, intervenes in the mature stages of growth between 

 the tabulse and the wall." (22, 158.) 



In this definition he includes species of the type of C. comu-hovis, 

 which he considered identical with C. comucopicB. 



The most recent paper on Caninia is that of Achille Salee, ^^Contribu- 

 tion a FEtude des Polypiers du Calcaire Carbonifere de la Belgique," 

 published in 1910. Much of his discussion is based on Carruthers' w^ork 

 and he gives no synonymy, merely referring to that given by Carruthers. 

 His paper, however, is comprehensive and brings out many points not 

 previously emphasized. He states the differences between Caninia, in 



