i ig i i a i eel ee ee 
C.—GEOLOGY. 93 
Practically all the graptolites referred to above, which are the pre- 
dominating element in the fauna, are characterised by simple cells— 
i.e. at most a reproduction of the embryonic sicula slightly modified in 
respect of relative length and breadth—and they follow what I have 
called elsewhere the Dichograptus plan of development; they may, 
therefore, be regarded as constituting the first or Dichograptid Fauna, 
which is pre-eminently characteristic of the rocks of Arenigian or 
Llanvirnian age. Without any special knowledge of species or genera, 
the horizon of this fauna may be recognised by the presence of branched 
graptolites with simple thece, the presence of scandent forms being 
indicative of the higher beds. 
In all the earlier graptolites, as has been shown, the cell type is 
simple, but soon after the two-stiped horizontal Didymograpti have 
developed a slight change begins to be apparent in the thece of some 
forms; this shows itself in a drawn-out curvature of the cell wall and a 
turning in of the apertural margin, which gives a most striking and 
characteristic appearance to the cell after compression. _This is first 
apparent in the thece in the region of the sicula, and becomes less 
conspicuous as the stipe grows in length; for it may be noted at this 
point that all progressive development (anagenesis) is first indicated in 
the proximal and, therefore, youthful region of the rhabdosoma, and 
when retrogression (catagenesis) occurs, it is in this same proximal 
region that signs of former elaboration are retained. 
Throughout the earlier rocks of Llandilian age the great majority 
of the graptolites have cells of this slightly elaborated type and two 
stipes only, which are reclined or reflexed in their position of growth ; 
but gradually in some forms an increasing degree of curvature of the 
walls of the cells becomes apparent, and the incurving of the apertural 
region is accompanied by a degree of torsion that after compression 
causes a very different appearance according to whether the rhabdosoma 
is viewed from the front (obverse) or back (reverse). This is the 
Dicellograptus stage, and so distinct is the appearance of this graptolite 
from any Didymograptus that it would never be considered related if the 
successive stages had not been followed step by step. It may be noted, 
too, that whilst this cell elaboration is in progress evolution along other 
lines seems to be temporarily arrested, but is resumed when the elabora- 
tion has reached its acme, especially towards the attainment of the 
scandent position of growth; this is at first only partial, as in Dicrano- 
graptus, but is eventually complete, as in the closely related Clima- 
cograpti, which are scandent throughout. The relationship of 
Climacograptus is clearly with Dicranograptus and Dicellograptus rather 
than with Diplograptus, with which up to the present it has been invari- 
ably grouped. The only connection it really has with Diplograptus is 
that, being a biserial scandent form, it is at a similar evolutional staqe. 
Since the simpler type of thecal elaboration is characteristic of the 
graptolite Leptoqraptus, the various forms in which this type of theca 
is found may suitably be regarded as constituting the second or Lrpto- 
GRAPTID Fauna, and its occurrence, whether in simpler or more complex 
forms, may be taken as indicating rocks of Llandilian or Caradocian 
age. As will be shown later, other features more particularly charac- 
’ 
