xrn PHYLUM CHORDATA 617 



off the direct line of ancestry of other species, but, as far as we can 

 judge, the same is true of most genera and families, of a large 

 majority of orders and classes, and even of most phyla. It would 

 certainly seem that existing Chordata, Mollusca, Arthropoda, 

 Annulata, Echinodermata, Nemathelminthes, and Platyhelminthes, 

 are all independent branches of the animal tree, having no con- 

 nection with one another save through the trunk. 



There are, however, existing groups which seem to represent 

 actual stages in the existence of others. For instance, it can 

 hardly be doubted that Amphibia are derived from Fishes and 

 Birds from Reptiles; that if we could discover the unknown 

 ancestors of those classes they would be classed definitely among 

 Pisces and Reptilia respectively, though probably not belonging to 

 any known order. 



In the same way everything seems to point to the conclusion 

 that all the higher phyla must have passed through some kind of 

 coeleuterate stage, and, before that, some kind of protozoan stage, 

 so that these two phyla may be said to represent actual steps 

 in the evolution of the higher forms. It is therefore legitimate to 

 assume, in the absence of direct evidence, that the ancestors of 

 both the Coelenterata and the Porifera were unicellular or "non- 

 cellular " forms, i.e., to be classed among the Protozoa, and that the 

 ancestors of the nine higher or triploblastic phyla were diploblastic 

 forms, i.e., to be classed among the Ccelenterata. 



Most, if not all, of the triploblastic phyla appear to be terminal 

 or culminating groups. There is no reason for thinking that either 

 of the three highest phyla — the Chordata, Mollusca, and Arthropoda 

 — ever passed through a stage which, if known, would be classed 

 among Platyhelminthes, Nemathelminthes, Echinodermata, or 

 Annulata, The wide occurrence of the trochophore, or some 

 similar larval form, seems, however, to indicate a certain bond of 

 union. The typical trochophore of Annulata and Mollusca, the 

 echinopsedium of Echinoderms, the ciliated larva of Molluscoida, 

 the tornaria of Balanoglossus, and the adult Rotifer, present, amid 

 endless diversity in detail, common characters which, in the absence 

 of better evidence, may be considered as indications of remote 

 affinity. The Arthropoda alone among the higher phyla are devoid 

 of even this slender connection with lower forms : there is no indi- 

 cation throughout the phylum of anything approaching to a 

 trochophore ; the crustacean nauplius is quite sui generis ; and the 

 larval forms of Insects and Arachnids simply suggest a 

 homonomously segmented ancestor. This suggestion is supported 

 by Peripatus, the cilia, hollow appendages, nephridia, and ladder- 

 like nervous system of which seem to point to its derivation 

 from a segmented " worm " not far removed from the annulate 

 type. 



