138 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



be attended to — namely, the character of the cellnles, and the nature 

 of the base. The last of these is in most cases by far the most im- 

 portant, since two distinct species may exactly resemble each other 

 as regards the shape of the cellules, and may yet be easily separated by 

 examining the processes at the base, as is the case with 2). teretiusculus, 

 His., and D. hicornis, Hall. This character, however, has been far 

 too much neglected ; and though seldom of itself sufficient for specific 

 determination, it affords, when taken in conjunction with the cellules 

 and general shape of the frond, the most valuable aid to a correct 

 diagnosis. So much is this the case that the description of hardly 

 any species of Di^plograpsus can be considered absolutely free from 

 doubt, unless the nature of the base be distinctly specified. On this 

 view of the importance of the base as a specific character, the species of 

 Diplograpsus may be conveniently grouped into three main sections. 

 In the &st of these the base is characterized by a median radicle, 

 sometimes very rudimentary, flanked by two lateral processes, of 

 varying length, which spring from the primary cellules on each side, 

 as is seen in I), hicornis, Hall, D. pHstis, His., D. antennamus, Hall, 

 D. Whitfieldii, Hall, &c. In the second group the two primary 

 cellules are greatly elongated, and form "with the solid axis a broad 

 tapering radicle, as in D. cometa, Gein., D. pcdmeus, Barr., and 

 D. acuminatus, Nich. In these, however, no true " radicle " exists, 

 since the solid axis is not prolonged below the first cellules. In the 

 third group the base is formed by a basal extension of the sohd axis, 

 sometimes to a very great length, beyond the proximal extremity of 

 the frond, as in I), teretiuscuhis, His., and D. pristiniformis, Hall. 

 It is hardly necessary to remark that this must be carefully dis- 

 tinguished from the extension of the axis beyond the distal extremity 

 of the frond, a character common to almost all (probably to all) the 

 known species of Diplograpsus. 



It is not pretended that these groups are strictly natural ; but they 

 embody the three main types displayed by the base of the Diplograpsi, 

 and they afford a convenient means of separating the numerous and 

 ill-defined species of the genus. It is quite possible, too, that all 

 the known Diplograpsi cannot be referred to any of these three 

 sections ; and we know so little of the real significance of the basal 

 processes, that there is some reason to believe that even a single 

 species may at one period of its grovrth belong to one group, and at 

 a later period to another. Thus I may mention that I have found 

 in the Upper Llandeilo rocks of Dumfriesshire specimens of a Diplo- 

 grapsus undistinguishable in other respects from the ordinary form 

 of D. teretiuscuhis of Hisinger, but provided with two long lateral 

 processes in addition to the basal extension of the axis. This may 

 possibly be a new species, but may with equal probability be merely 

 a particular stage of the development oiD. teretiusculus, His. 



The first and third of these groups are alone represented in the 

 Skiddaw Slates, from which four species of the genus are now known 

 to me — namely, D. antennarius, Hall, D.mucronatus, Hall,i>. teretius- 

 culus, His., and D. pristiniformis, Hall, of which the first and last are 

 confined to the Quebec Group. 



