140 BRITISH PALZZOZOIC FOSSILS. [Crustacra. 
of their structure. The apparently anomalous nature of the facial suture has been spoken of by Burmeister, 
who saw no clue to its nature; the nature of the parts outside the eye-line, or ‘ wings,’ as they were called, 
has also been alluded to by authors as inexplicable ; while those who, comparing the Trilobites with Branchipus, 
supposed the body of the animal to occupy the axal lobe only, have expressed their astonishment at the 
eyes being placed on the lateral lobes, or ‘cheeks.’ When we bear in mind that the carapace of a crab, 
for instance, is a great backward prolongation of one of the rings of the head, and is quite distinct from 
the posterior cephalic and the thoracic segments which it covers, and which exist in a membraneous state 
beneath it, we are prepared to admit, that though the segmental furrows on the glabella of many Trilobites 
indicate cephalic rings, they by no means prove the cephalic shield to be formed of the anchylosis of such 
rings, which may only exist below, impressing the shield like the various regions on the back of a crab. To 
determine of what rings the shield is composed, I started with the main characteristic of the first ring of all 
crustacea, which is, to bear the eyes when they are present ; the second and third rings bear the antenne; and 
the remainder of the cephalic rings bear the parts of the mouth. The eyes of Trilobites, when they exist, are 
always connected with the piece anterior and external to the eye-line; this piece is usually continuous 
from side to side at the front margin, and I think is probably the first or ophthalmic ring; its lateral 
portions produced backwards, and bearing its peculiar appendages, the eyes, with it: every ring being 
theoretically divisible into six pieces, affords an explanation of the suture which sometimes separates the 
two parts in front, and even of the rostral shield of Calymene (if it belongs to this ring). On this view 
the facial suture becomes at once intelligible as the line of separation between the first and second cephalic 
rings, analogous to the divisional line between one thoracie ring and another. The piece within and behind 
the eye-line should on this supposition be the second or antennary ring ; and as remarkably supporting this, 
I must refer to p. 42 of my ‘Synopsis of the Silurian Fossils of Ireland,” where I announced the discovery 
of the remains of antenne, as a deep pore on each side of the front of the glabella, in the furrow which 
surrounds it, and in which, when clear of matrix, I have observed them in Trinucleus, Acidaspis, Calymene, 
Ampyx, Griffithides, &e. We would thus have the cephalic shield of Trilobites composed of an extension 
of the first two cephalic rings. The protuberance called the glabella probably contains the stomach, which is 
always in crustacea large and placed over the mouth ; the segmental furrows on the sides, indicating the rings 
beneath, which bear the parts of the mouth, 
The arrangements hitherto proposed of this group fall far short of what could be desired for a satis- 
factory or natural classification. I have satisfied myself that the number of the segments, so far from 
being of family importance, as hitherto supposed by some authorities, cannot indicate more than subgeneric 
differences among Trilobites, being an instance of a law well known to every experienced zoologist, that 
those characters of most vital importance in any class or order are often found suddenly to lose their 
importance in some subordinate section of the group. Thus in the whole class Crustacea the number of 
segments is of great importance, and scarcely varies in the largest groups, while in the Hntomostraca alone 
(to which the Trilobites belong) we have every variety, from the smallest to greatly the largest number 
of segments known in the class. The consideration therefore of the characters of all the recent and fossil 
Crustacea, taken as a whole, warns us against laying the stress, which has hitherto been done, on the number 
of segments as a general character for the formation of the great groups. Such a use of this character 
separates the most nearly allied forms into distinct families, and allies those agreeing in no other points. 
The two chief systematic arrangements which have been proposed are those of Dr Burmeister, taking as the 
principle of division the presence or absence of such contractile power as enables the animal to roll itself into 
a ball; and that of Hawle and Corda, who rest their great divisions on the margin of the pygidium 
being toothed or entire. ‘The latter character seems to me of no more than specific importance, and the 
former, though of imperfect application and not ascertainable in many cases, is included in the following 
mode of arrangement, which, after a careful study of the British and Foreign types, I venture to propose, 
founded on a consideration, in the first instance, of the structure of the plewrw, or lateral portions of the 
thoracic segments, which I find to afford definite characters easily ascertainable in all moderately well- 
preserved specimens. I propose to divide the Trilobites into the five following subfamilies ;—Ist, Agnostine ; 
2nd, Harpedine ; 3rd, Ogygine ; 4th, Paradoxine ; 5th, Asaphine. 
