400 MR. H. M. BERNARD ON THE 



the alimentary diverticula became specialized, tlie glauds underwent variations, their 

 ends often meeting and fusing. The reticulate form found in Scorpio arises from a 

 series of such fusings between the alimentary diverticula ; it is certainly not primitive. 



The reticulate form of the sexual gland iu Limulus is more highly speciaHzed thau was that of the 

 primitive Arachnid. 



The ^Resemblance of Scorpio to the Eiirypterids, due to Convergence. — The foregoing 

 arguments show very conclusively that there can be no real relationship between the 

 Merostomata and the Arachnids. The great resemblance between the Eurypterids and 

 Scorpio must therefore be one of convergence. This is not difficult to demonstrate. In 

 the first place, it is obviously so if Scorpio is a specialized and not a primitive form, and 

 this we have abundantly shown. Indeed, the very point in which Scorpio most 

 resembles the Eurypterids — namely, the great compression of the anterior limb-bearing 

 end of the bod}' — hapj)ens to be its most marked, specialization, which has caused it to 

 depart furthest from the primitive condition witli free segments. 



But the argument can be made very much stronger. It is obvious that if the resem- 

 blance between the Eitrypterids and Scorpio is not a case of convergence, then, as we 

 go back along the lines of their development, we should find this resemblance getting 

 more and more striking, until we reached their common ancestral form. In order thus 

 to trace the two back to their sources, the advocates of their relationship have made 

 a close study of their embryological development, and have very naturally discovered a 

 wonderful parallel. But embryology, in this case, teaches us nothing, because each form 

 is laid down from the first in its definitive shape, so that the likeness of the embryo is but 

 a repetition of the likeness of the adult. But where embryology fails us, comparative 

 anatomy comes to our aid ; its study will, I expect, be found to reveal the phylogenetic 

 history of animals and groups of animals with more precision than embryology 

 ever can. This is certainly so in the present case, in which embryology does not in 

 either case teach us anything clear as to what its less differentiated ancestors were like. 

 Thanks to comparative anatomy, we can reconstruct iu some important details the 

 ancestors of the Arachnids before the segmentation became too specialized ; we can 

 ascertain the leading specializations of that segmentation, marking off for all time the 

 Arachnid phylum from all other derivatives of the Chsetopod Annelids. Further, 

 Arachnids still persist in which only three segments are fused together, all the rest being 

 free. Turning now to the Merostomata, — do we know what their earlier, less differ- 

 entiated forms were like — those forms, for instance, in which only the three anterior head- 

 segments were fused together ? If so, does it resemble the Arachnid at that same stage 

 of segmental differentiation more closely than Slimonia resembles Scorpio ? No, the very 

 reverse is the case. We can deduce the Eurypterids with some precision from less 

 specialized, more richly segmented Trilobite-like forms with head-segments in all stages 

 of fusion — five being the usual number. But all resemblance to the Arachnids has gone, 

 whether to those which we claim to be specialized or to those which, having only three 

 segments fused anteriorly, we claim to be less specialized. We thus find that, as we trace 

 the two forms back to their less differentiated ancestors, all resemblance between them 



