STUDY OF THE GRAPTOLITES. 235 



sometimes absent, apparently from accident, and some of the four-stiped 

 species are not known to possess it ; while it has never been observed in 

 any of the species where the stipes are properly branched, or divided 

 in the celluliferous parts of their length* 



The interior of this corneous disc, previously described as apparently 

 composed of two plates of the test, has probably been occupied by some 

 softer substance, which may have been an extension of the common 

 body, or have possessed in some degree the character of the common body 

 of the stipe. 



The development of the diprionidian forms, as deduced from the young 

 Graptolites which we find associated with the reproductive cell-bearing 

 stipes, would show that these forms of Graptolites exist as single and 

 simple independent stipes from the commencement of their growth. 

 Nevertheless, I conceive that both Retiolites and Retiograptus may 

 have existed in compound fronds, having their origin from a central 

 point not unlike in the commencement to Gvaptolitlms logani, but without 

 the central disc. These fronds were probably concavo-convex, as 

 were the individual stipes. The solid axis, instead of being central, 

 is placed externally along the centre of the convex or outer side ; and 

 the cell-divisions on that side proceed from it; while on the upper 

 or concave side the cell-divisions do not reach the centre, leaving a space 

 for the common body, which has been shown by Barrande and Geinitz 

 to produce a central longitudinal prominence. 



In these forms the mode of development has been similar to that 

 already explained, the modification being chiefly in the external position 

 of the axis and the joining of cell-divisions with the axis on one side ; 

 leaving the common body in a somewhat triangular form, from which 

 the alveoles are developed on either side. 



Whether the Phyllograptus existed as simple stipes with four ranges 

 of cellules, or in a compound arrangement as in Retiograptus, the mode 

 development has been similar: either the germ with its initial point 

 developed a single stipe with four ranges of cellules, or the same elements 

 first subdivided, and each division gave origin to its stipe through the 

 common body. 



In regard to the development of the cellules in the different parts of 

 the graptolite, we observe, as a uniform feature, their smaller size 

 towards the base of the stipe. In all the monoprionidian forms, this 



* See remarks upon the Genus Dichogbaptus under Generic descriptions in this paper. 



