78 SYSTEMATIC ARRANGEMENT 



3. The fact that the size and shape of the caudal shield correspond pretty nearly with 

 the size and shape of the cephalic shield. This is no doubt the case also in Ogyyki, and 

 partially in Trinudeus, but less generally than in the present group. We shall rarely 

 meet with forms in which the caudal shield is much smaller than the cephalic shield; and, 

 on the other hand, we shall find some species in which the former is the greater of the two. 

 The size of the caudal shield can scarcely therefore afford any safe inference with regard 

 to the power of the animal to roll itself up. 



Since most of the members of this group are found in limestones, they appear, upon 

 the whole, to be of less ancient date than the members of the previous group ; the oldest 

 forms are those from the limestones of the lower Silurian system, as are also the genera 

 Amjji/x, Asaphim, and lUcenus, taking them in their correct limits ; some peculiar species are 

 also said to have been found in Tafel-schiefer. 



Calymene and Fhacops follow next, but the different species range through all the 

 stages of the Silurian formation. Certain genera furnished with nine or ten body rings and 

 a highly convex, axis, are, however, decidedly more modei'n, and seem to form the last link 

 in the series of creation of these forms. With regard to the division of this very natural 

 group into subordinate genera there appears to be one fact which has hitherto always been 

 overlooked ; namely, the nature of the shell. I have already alluded to this, and have 

 shown, when treating of general considerations, that many Trilobites which now appear to 

 us to be smooth, were furnished during life with a peculiar granulated layer ; that others, 

 on the other hand, had a peculiar sculpture and punctation in the shell itself, which is 

 incompatible with the presence of a special upper membrane. Pursuing this idea, we 

 discover that Trilobites with distinct granulations always possess a more than ten-jointed 

 axis of the body, which in that case contracts very much posteriorly ; the Trilobites 

 without granulation, on the other hand, never have more, and usually exhibit less than ten 

 joints, of equal width in the body. Hence I believe myself justified in assuming that all 

 Trilobites, furnished with an axis of ten joints which gradually becomes smaller towards 

 the posterior part, possessed a granulated upper membrane, while those furnished with a 

 fewer number of joints possessed a truly punctated or sculptured shell. There are, however, 

 in both groups, genera in which the number of joints is ten. 



This difference becomes yet more marked when we consider that the Trilolsitcs not 

 granvdated occur only in the most ancient and the newest Palfeozoic strata, while those, on 

 the other hand, which are granulated, are found in the middle beds of that series. I have 

 employed this structure as a permanent principle of classification,* and thus form two 

 subdivisions of Trilobites having the power of rolling themselves up. 



* Dr. Beyrich lias questioned the value of this distinction, because in the genus Bronteus some 

 species have a granulated, and others a lineated surface of the shell. But this genus belongs to the 

 group of Trilobites not rolling themselves up, and does not therefore affect the question with regard 

 to the other group in which we make use of the principle. A more important objection would appear 

 to be, that certain species of Archegonus or PhilHpsia are granulated, and others lineated, since these 

 genera do roll themselves up. It seems that in this genus, the last effort of a once numerous group, 

 the character in question has degenerated into a mere specific distinction, although once distinctive of 

 the main divisions of the genera, 



