STUDY OF THE GRAPTOLITES. 247 



covered by a net-work of slightly different texture from that of the other 

 portions of the substance. 



The Canadian specimens which I have referred to this genus are so 

 extremely attenuated that it is impossible to determine the details of 

 structure, and the surface-characters are obscure. Thus far we have no 

 American specimens in a condition to aftbrd the means of elucidating 

 some obscurities, which seem to me still to exist, in regard to the inti- 

 mate structure of this genus. The species of the Clinton formation is 

 extremely compressed ; and while some specimens show the cell-divisions 

 terminating at a distance from the centre, yet, after protracted and 

 repeated examinations, my most critical observations result in showing 

 only the structure which is illustrated on Plate i, figs. 20 and 21. 



On one side we have an external, cylindrical, solid axis, to which the 

 cell-divisions are joined ; but these latter show only a filiform cylindroid 

 process, extending ii-om the axis to the cell-mai'gin, and projecting a 

 little beyond the margin of the stipe. The only other aspect which we 

 observe in this species is that of an undulating or zigzag filiform axis on 

 the opposite side, to which the cell-partitions are joined, as in Plate i, fig. 

 21. We know this to be on the opposite side or within the stipe, as it 

 is sometimes seen overlying the straight axis and cell-partitions. 



At the junction of the cell-divisions with the zigzag axis there are 

 other processes of similar character, projecting upward and outward from 

 the axis, all of nearly equal length, but apparently broken at their 

 extremities. I have not been able to determine any connection between 

 these and other parts of the skeleton, but we have the two structures 

 very clearly represented in the figures referred to. I have supposed that 

 similar processes may have extended to the opposite side, from the junc- 

 tion of the cell-partitions with this undulating axis, either joining the 

 cell divisions or the straight axis; but after long investigation, I have 

 been unable to find satisfactory evidence of such connection. The cell- 

 apertures are surrounded by thickened margins, and the only appearances 

 of cell-partitions are the sub-external cylindrical extensions from the 

 aperture to the axis. Neither the species of the Quebec group nor that 

 of the Clinton formation, in any of the specimens seen, are in a condi- 

 tion to show evidence of the concavo-convex character of the stipe 

 represented by M. Barrande and Prof Geinitz. 



The species of Retiograptus, while having some characters in common 

 with Retiolites, do not possess the reticulated structure of the test in 

 either of the described species. There yet remains some obscurity 



