STUDY OF THE GRAPTOLITES. 245 



allied forms, when flattened in the shale, show, as already described, a 

 simple semi-elliptical notch in the margin of the stipe, nearly rectangu- 

 lar to the axis. This is well shown in fig. 3, Plate vi, of M. Baebande's 

 memoir, and also in Mr. Salter's illustrations of Graptolitlius teretiuscidus 

 of HisiNGER.* It is represented, lest perfectly, in the figures of Prof 

 Harkness,! and in most of my own figures on Plate Ixxiii of the first 

 volumes of the Palaeontology of New York. When compressed rectangu- 

 larly to the cellules, the apertures are transversely oblong-oval; and the 

 same form is shown when looking upon the celluliferous margin of an 

 uncompressed stipe. 



The structure of these stipes ^and their cellules has already been 

 described in a preceding section, with reference to the figures illustrating 

 the same. The G. hicornis, known in New- York and Canada, may be 

 considered the type of a group of species of which we have two in the 

 shales of Norman's Kill near Albany, one in Ohio, and a similar or iden- 

 tical form in the Utica slate at Collingwood in Canada West. I would 

 include in the same group figs. 5 and 6 of Plate iii, as well as figs. 7, 8 

 and 15, Plate ii, of M. Barrande's Memoir, Graptolitlius teretimculus of 

 HisiNGER, and those referred to the same species by Salter.^ The 

 Diplogrdptus rectangularis of M'CoY§ is of the same type, as also figs. 

 1, 5, 10, 11, 12, etc.. Tab. ii of Geinitz {Graptolithen) ; and I con- 

 ceive that many of the scalariform specimens belong to species of this 

 character. 



The Gmptolithus ramosus has usually been arranged by authors under 

 DiPLOGRAPTUS ; the lower part of the stipe being simple and having a 

 double range of cellules, while above the bifurcation it has a single 

 range on the outer margin of each division, as already described ; and 

 a simple explanation of this condition has been offered by supposing that 

 the solid axis has been divided after the death of the zoophyte. This, 

 however, will scarcely afford a satisfactory argument when we find that 

 all the specimens are in the same condition ; that usually the division 

 begins at a uniform distance from the base ; and that when entire, the 

 divided portion much exceeds the simple undivided part of the stipe. 

 Moreover, the species is recognized in this condition in the Hudson-river 

 formation in Canada, and has likewise been recognized in Great Britain ; 



* Quarterly Journal of the Geological Society of London, Vol. viii, PI. xxi, flgs. 3 and 4. 



t Quarterly Journal of the Geological Society, Vol. vii, PI. i, fig. 11. 



J lb. ibid., Vol. viii, PI. xxi. 



§ British Palaozoic Fossils, page 8, PI. xiii, figs. 8, 9 and 10. 



