3) ile CANADIAN FOSSILS. 
——s 
River formation at Marsouin, Canada, the stipes and cellules are less - 
fully developed than in those of the same species from Norman’s Kill near 
Albany, while the mucronate extensions from the cell-apertures are more 
conspicuous. 
Besides these ornaments, there is on each side of the radicle or initial 
point at the base of most of the diprionidian species of graptolites, a 
small process, varying in length, and usually directed downwards. These 
processes are usually short, but often considerably extended; in some spe- 
cies they are very slender, while in others they are strong and rigid. 
In G. pristis they are frequently seen as short slender processes; while 
in G. bicornis they are rigid, strong, and slightly curving. In G. anten- 
narius, a congener of the latter, they are long and slender setiform 
appendages. In one species of Retiograptus they are slender setiform 
processes, directed downwards. 
In no species of Phyllograptus have such appendages been observed ; 
nor have they been seen at the bases of the stipes of etiolites. 
§ IV.—MODE OF REPRODUCTION AND DEVELOPMENT IN THE GRAPTOLITIDA. 
As already remarked, the graptolites proper are now generally referred 
by authors to the Radiata; while some forms which I include in the 
family have been heretofore regarded as reticulate bryozoans, or as gor- 
gonians. ‘The nearest analogues in the recent fauna appear to be among 
the group Pennatulide, or in the Sertularide ; but in all these there is 
no absolute identity in the mode of development or character of cellules, 
so far as my observation has extended. 
In nearly all the true bryozoan forms among fossils, we have the means 
of tracing the relations and analogies, both in manner of growth and 
reproduction, throughout all the successive geological periods, and in the 
present fauna. It becomes therefore more difficult to discover such 
analogies for the Graptolitide, since the graptolites proper disappear 
from existence in the Silurian period; and the latest form of Graptolitidee 
(Dictyonema) is not found, so far as now known to me, in American 
strata, at a later period than the Hamilton formation or Middle Devonian. 
From this cause, the mode of growth and development are not so readily 
understood as in those families which can be traced throughout the geo- 
logical series, and still find their analogues in the present seas. 
In 1858, I laid before the American Association for the Advancement 
of Science a notice, with some illustrations of graptolite stipes, bearing 
what I then regarded, and do still regard, as the reproductive cells or 
ovarian vesicles. These cells first appear as small ovate buds upon the 
