C.—GEOLOGY. 105 
C. parvula of D.d.2. is at a stage of development similar to that of 
C. cambrensis, as is also C. pulchra at the same horizon in yet another 
species-group. 
It would, I believe, be perfectly possible to adopt a classification of 
the whole of the Ordovician based upon the evolutional sequence of the 
various Calymenes. 
All these facts illustrate that even from the purely paleontological 
standpoint much field knowledge is essential if a right conception is to 
be gained of the true relationship existing between species and species. 
It appears to be also in the highest degree necessary to view a succes- 
sion of forms like those I have quoted in order to determine what 
characters are really of importance in the recognition of species. 
Also, when lines of evolution result in the attainment of successful 
forms, not only do these appear to be numerically abundant, but it 
would seem also that they have a wide distribution in space. 
So much, then, for a possible line of evolution in the head of a, 
trilobite; we may next consider the evolution of the pygidium in a 
very different form. An interesting study of this appears to be afforded | 
by the species-group of Encrinurus punctatus. As is well known in 
the commonest type of this trilobite occurring in the Wenlock Lime- 
stone of Dudley, the axis of the pygidium’ shows a far greater degree 
of segmentation than do the lateral lobes; this may be interpreted 
as implying that numerous segments have been incorporated into the 
tail with a greater degree of fusion in the side lobes than in the axis. 
The species, moreover, is commonly recognised as possessing two well- 
marked varieties, var. arenaceus and var. calcareus, differing chiefly 
from each other in the possession of a definite mucro in var. calcareus, 
which has been interpreted as being connected with the supply of 
calcareous matter available, but, viewing the species-group as a whole, 
it would seem rather to be the natural culmination or acme of a definite 
tendency to fusion which is developed with increasing persistence 
throughout its history in time so far as I have been able to study it. 
The earliest forms which I have examined are to be found at the 
horizon of the Stinchar Limestone in Scotland and the Derfel Lime- 
stone of Wales. The graptolite shales associated with these limestones 
prove their age to be Llandilian. At this horizon the relation between 
the segments of the axis and the lateral lobes of the pygidium never 
exceeds 2:1, whilst in the two earliest segments the proportion is very 
clearly 1:1; in the Caradocian the proportion rises to 3:1 for segments 
5 and 6, whilst the Ashgillian forms (multisegmentatus stage) show 
2:1 for segment 2 and still 3:1 for segments 5 and 6. In the Lower 
Valentian segment 4 has risen to 3:1, whilst in the Upper 
Valentian it is commonly the third, though there is some variation, 
since in some cases all that it is possible to make out is that there are 
five segments in the axis compared with two (2 and 3) in the lateral 
lobes. In the succeeding Wenlock forms the culmination is reached with 
3:1 for segments 2-5, and 4:1 at the sixth; in all these later forms 
there is a tendency to fusion of the later lateral lobes with the axis, 
partially, as in the case of 7 and 8, throughout their length, and more 
definitely at their terminations. 
