F. R. Cowper Reed — On the Phacopidce. 175 



ihe whole history of the group must be undoubtedly interpreted as 

 ■of considerable importance, though even in the larval stages some 

 tachy genetic features appear which stamp the group as belonging to 

 one of the higher and later families of trilobites. There does not 

 seem to be any reason for considering the pentamerism of the later 

 a,nd more specialised forms of Dalmanites as signs of retrogression or 

 as senile re-adoptions of pre-adult and early characters ; for the 

 succession of Dalmanites-f orms is traceable in practically a con- 

 tinuous and unbroken series from their first appearance to their 

 extinction. 



The loss of the original pentamerism of the head-shield to 

 a greater or less extent, as illustrated by a large number of members 

 of the family, must be taken as a departure from primitive conditions 

 and as a sign of specialisation and an advance in development. 



This structural change may take place in either of two ways, 

 (1) by the fusion of two or more of the successive lobes; or (2) by 

 the unequal growth of the lobes, leading to the reduction in size or 

 squeezing-out of some of them. In the first case the obsolescence 

 or complete disappearance of the segmental (lateral) furrows 

 ■concerned accompanies the change ; and in the second case the 

 lobes themselves are more or less lost. 



It is noticeable that these types of structure in the head-shield do 

 not absolutely replace each other in regular chronological succession, 

 but frequently exist contemporaneously ; the more highly specialised 

 type did not drive out the simpler type, and the old primitive stock 

 from which the others must have been derived held its ground in 

 spite of the vigorous growth of its lateral offshoots, even lasting till 

 the final disappearance of the family. 



A lai'ge number of intermediate conditions between the perfect 

 pentamerous. and the lobeless or paucilobed forms of Phacopidae 

 have been observed, and this has given rise to the abundant and 

 varied nomenclature. Many of the so-called genera, subgenera, and 

 groups or sections have been founded on these transitional forms, 

 which are of much interest from a developmental point of view, but 

 these distinctive names of higher than specific rank have obscured 

 the broad features of the phylogeny, particularly in cases where the 

 occurrence of such forms is purely local and their representation 

 i:estricted to a single species. 



The presence of so many intermediate links has rendered the 

 division of the family into sharply separated and well-defined 

 genera a matter of extraordinary difficulty, as the literature on the 

 subject illustrates, hardly two authorities agreeing completely on 

 the limits of the various divisions of the family. 



Lines of Development. 



Eegarded as a whole we may distinguish three lines along 

 which the development of the Phacopidas has proceeded, and they 

 start from an early stage in the history of the family. Two of 

 these lines find their culmination in the Devonian period, where 

 they are represented by the typical FJiacops with a lobeless or 



