female genital appendage [text fig. 19]. A like pair of tubes has been found by the writers in E. remipes and lacustris [pl. 12, fig. 2]. Finally, the delicate membranes of the interior of the ventral side and the oval attachment areas of the gill plates were described and reproduced by photographs.
It so happens that E. fischeri, which thus has become known down to the least details of its integumental structure, differs very little from the genotype of Eurypterus, E. remipes; and both Schmidt and Holm have suggested that these with E. lacustris may be only geographical varieties of the same species, inasmuch as they appear at the same geological horizon. But the question remains open whether E. remipes, the genotype by right of priority, is a faithful expression of the structure displayed by the genus Eurypterus. To determine this point we may briefly survey the phylogenetic relations of the species of Eurypterus.
The American species of Eurypterus readily fall into three subdivisions, viz, (a) those that group themselves around E. remipes; (b) those that vary in different directions from this group but are approximately contemporary; and (c) the later Carbonic species. Group (a) is represented by E. remipes, E. lacustris and its var. pachychirus, E. pittsfordensis, and possibly also E. maria. It is Eurypterus sensu stricto. Group (b) is a diffuse association comprising aberrant species E. dekayi, E. microphthalmus, E. pustulosus, E. kokomoensis, E. prominens and E. boylei. Eurypterus microphthalmus is representative of a small group, also present in England (E. brewsteri), characterized by its small compound eyes and broad, short cephalothorax. Eurypterus dekayi is probably a branch from E. lacustris with a number of characters already suggestive of approach to a phylogerontic condition. Eurypterus pustulosus, although attaining immense size, is aberrant in the exaggerated development of the tuberculation. Eurypterus kokomoensis stands apart in a number of char-
