Grajptolites. 287 



vidua! polypite from the common ccenosarc. M. Barrande 

 describes an internal orifice as existing in this position, but his 

 longitudinal section of G. priodon gives no indication of it, 

 and his figures illustrating it in G. nuntius, Barr., and G. Halli, 

 Barr., show that he has been misled by referring to these 

 species scalariform impressions of a double graptolite. His 

 internal orifices are the external openings of the under surface, 

 showing themselves faintly through the chitinous investment 

 of the polypary, while the external orifices are the series of 

 openings preserved on the upper surface, which are conse- 

 quently preserved with greater definition. 



The back of the polypary was strengthened by a solid axis 

 (Fig. 2) composed of the same substance as its walls. This is 

 a distinct structure, and capable of being separated from the 

 graptolite without destroying that surface to which it is closely 

 applied. It is frequently continued beyond the cell-bearing 

 portion of the graptolite as a simple naked axis. 



The structure of G. Roemeri (Fig. 16) agrees in every 

 respect with that of G. priodon, except that the walls of the 

 neighbouring hydrothecae are in contact throughout their 

 whole length. In this respect there is a considerable variation 

 among graptolites. In G. convolutus, His. (Fig. 14), the 

 hydrothecae are free throughout their whole length, while in 

 Barrande's genus, Rastrites (Fig. 17), the individual hydro- 

 thecae are not only free, but separated from each other by an 

 interposed non-polypiferous portion of the filiform polypary. A 

 very different structure appears to exist in G. latus, M'Coy, 

 and G. Sagittarius, His. Both these species have as yet, I 

 believe, been found only in a flattened condition, so that the 

 full meaning of the external appearances cannot be positively 

 determined. In well-preserved specimens of both, there 

 seems to be a septum at the base of each hydrotheca ; and in 

 G. Sagittarius, I have noticed the divisions between the indi- 

 vidual polypites passing down to the solid axis, leaving appa- 

 rently no space for the common ccenosarc. The structure of 

 the polypary may, however, have been similar to what will 

 be presently described as existing in Hall's genus Clima- 

 cograptus. 



The double graptolite, for which M'Coy proposed the name 

 Diplograpsas, appears to differ little, if at all, from the struc- 

 ture of the single forms. We shall have a correct impression 

 of D. pristis, His. (Fig. 18), if we consider it as composed of 

 two specimens of G. Roemeri, united back to back, having 

 similar hydrothecae, but with the posterior portions of the 

 single polyparies removed, which, if present, would separate 

 the animals of the two series, and the solid axis remaining in 

 the centre of the common fleshy basis of the colony. Bichter 



