i^oAoru. NEW CHINESE PALEOZOIC FOSSILS— GIRTY. 39 



in trunsvorse sections, they are rather rare and their arrangement has 

 not been determined, 



An}^ statement as to septa depends upon the interpretation of certain 

 appearances in thin sections. The bounding* walls are not thick and 

 show a dense median line, which is usuall}^ more or less wavy. On 

 either side of the median line is a tolerably thick layer, intermediate 

 in transparency between the material filling- the intertabular spaces 

 and that of the median plate. The inner edge of this supplementary 

 wall is somewhat strongly denticulate, the projections being in some 

 places rather regular and in other portions of the same corallite al)sent 

 or irregular. Whether these projections arc really denticles or are 

 continuous ridges has not been ascertained. The ta))uhv show the 

 same degree of transparenc}' iis the supplementary wall, the projec- 

 tions of which can apparently with justification be interpreted as 

 pseudosepta, the dense median line l)eing the plane of cleavage sepa- 

 rating the truh' doulde- walled corallities. It is hardl}- possible to 

 count the pseudosepta or to ascertain the average number present. 



At first sight one might well feel somewhat doubtful where to 

 place this species, whether in ]:'}ivo,ntes or Mlckel'nwa^ the unusually 

 small size of the corallites certaiidy suggesting the former genus, as 

 well as the mural pores, which appear to be rather rare. Some sup- 

 port might also be found in the character of the tal>uh«, which, if 

 they are somewhat too convex, too irregular, and too confluent for 

 Fwrnsites^ possess these features in too slight a degree for typical 

 MicJn'lhwa. On the other hand, the somewhat fluted condition of 

 the median plate, which may be connected with the development of 

 pseudosepta, together with the apparent presence of well-developed 

 pseudosepta themselves, seem to distinguish it structurally from 

 FavositeH. The range of the latter genus, furthermore, appears to 

 terminate with horizons early in the Mississippian, while the present 

 form is of much later occurrence. Michelhiea^ on the other hand, is 

 already known in upper Carboniferous and Permian (?) terrains. On 

 this account it has seemed that the form under consideration should 

 be denied to Favosltes and placed with AUchellnea. 



Kayser figures an unidentitied species of Michelhiea from China 

 which difl'ers from the present one in the much greater size of the 

 corallites. By the same character Mlclielinea favositoides may be 

 distinguished from other mem])ers of the genus known to me, even 

 from the small-celled Kussian species M. coneinna Lonsdale. 



Local /fy and horizon. — Pennsylvanian (Wu-shan limestone); near 

 Ta-ning-hien, East Ssi-ch'uan (kStation 3). 



