Page:Walcott Cambrian Geology and Paleontology II.djvu/272



We now have from the Burgess shale specimens of a Middle Cambrian trilobite (Neolenus serratus (Rominger)), showing the character of its antennæ, legs, branchiæ, and caudal appendages, and another (Ptychoparia cordilleræ (Rominger)) with a full series of branchial appendages. Also there are five Ordovician forms, three from the Trenton limestone (Calymene senaria, Ceraurus pleurexanthemus, and Asaphus platycephalus) and two from the Utica shale (Triarthrus becki and Trinucleus concentricus), which preserve the antennæ, legs, and branchiæ.

A review of these shows a surprising uniformity of structure of the antennæ, legs, and branchiæ in genera separated by great intervals in the stratigraphic series, and distinguished by marked variations in the external form of the dorsal shield.

At first I was inclined to consider that the trilobite was a highly organized crustacean approaching the merostomes, but with the data now available I join with Burmeister and Bernard in considering that the trilobite is more closely related to the branchiopod crustaceans. Burmeister wrote in 1843:

The trilobites were a peculiar family of Crustacea, nearly allied to the existing Phyllopoda, approaching this latter family most nearly in its genus Branchipus, and forming a link connecting the Phyllopoda with the Pœcilopoda.

In order, however, to estimate fairly the affinity of the trilobites with the Phyllopoda, we must not lose sight of the important fact that the trilobites differ not only from the Phyllopoda, but from all other existing families of Crustacea in the varying numerical proportion of their thoracic rings; a peculiarity neither exhibited at present as a characteristic of any natural family among the Crustacea, nor in any of the heterogeneous Articulata. This peculiarity occurs, it is true, among the Aspidostraca (a group of the second great division of the crustaceans), but only in a modified form, the difference in the numerical proportion being always reducible to one fundamental number. This law is apparently not observed in the case of the trilobites.

It would seem then that the relation existing between the trilobites and the existing Crustacea is one rather of analogy than affinity, so that the whole group may be considered as a separate division, corresponding with the Aspidostraca in the formal variation presented from the typical character, but