GRAPTOLITES. 81 
an inch in diameter. Fragments of single stipes have been observed, 
having a length of six or seven inches, with a width not exceeding that given 
above. Surface usually smooth; the strize formed by the cell-partitions 
sometimes visible ; the back of the stipe somewhat thickened, and about 
one fourth of the width occupied by the common body. 
Cellules short and comparatively broad, very slightly curved ; about 
twenty-four in the space of an inch, and making with the axis an angle 
of about 40°. Margins of the aperture truncate, making an angle of 
98° with the axis: one third or more of the cellule is free; and near the 
radicle, one half of the length of the cellule is free. The partitions are 
distinctly visible in well-preserved specimens; but in most instances they 
are obscure. 
In this collection, the specimens are mostly upon weathered surfaces of 
the slate ; the substance is often partially removed, and no good impressions 
are preserved. This species is most nearly allied to G. s¢mzlzs, but is always 
more slender near the base: the serratures are more distinct in smaller 
stipes, and the angle of the cellules with the axis is much greater ; while the 
angle formed by the cell-aperture with the axis is much less than in that 
one. Ji differs equally from the allied Kuropean forms, in some respects 
approaching the G. sagittarius as represented by Geinitz (Graptolithen, 
-ete., pl. u, fig. 4); but it has more the aspect of G. Wilssoni (fig. 17 
of same plate). It differs essentially however from G. WVilssoni as repre- 
sented by Barrande (Graptolites de Bohéme, pl. 1, figs. 16 and 17), 
and from the figs. 19 and 31 of Geinitz, which our species never 
approaches in any stage of growth. In our specimens we see the con- 
nexion of the stipes with the radicle, and trace them continuously for four 
or five inches; and in separated fragments, we have specimens six or seven 
inches long. None of these offer an approximation in variations of form 
and proportional distance of cellules, to those represented by Geinitz 
as different phases of G. Milssont. We may add, that if such differences 
exist in the same species, we have then no means of fixing the limits of 
specific variation, or of determining the species among graptolites of this 
character, from the separated fragments. The comparatively large number 
of specimens in the Canadian collection affords good means of specific deter- 
mination for many of the forms; and unless we could find upon the same 
stipe, evidences of such variation, we should hesitate to consider the 
different varieties as one species. 
The G. Nilssont, as figured by Harkness (Quart. Jour. Geol. Soc., vol. 
. vil, p. 62, pl. i), differs from our species, and more nearly resembles the 
figures of Barrande. 
a 
