Mat 21, 1909] 



SCIENCE 



817 



to establish the tubular vertebrate nervous sys- 

 tem ; that the vertebrate gut is a new structure 

 formed by the fusion of the bases of arachnid 

 appendages; that the notochord was later 

 formed from this new gut; that the arachnid 

 genital ducts were retained as the vertebrate 

 thyroid; that the arachnid genital tissue and 

 liver became converted into the arachnoidal 

 fat which fills the brain-case and invades even 

 the skeletal capsule of the vertebrate ear ; that 

 Kupfier's sensory plakodes are reminiscent of 

 arthropod appendages; that coxal glands are 

 the homologues of both the pronephros and 

 thymus ; that the wide discrepancy in the plan 

 of the embryonic development of arthropod 

 and vertebrate, as in the inversion of the dorso- 

 ventral orientation, is a mere detail caused by 

 the shifting of the mass of the yolk. 



One hesitates at this day to reopen an in- 

 digestible discussion. And it would be profit- 

 less were it not that the volume has brought 

 forward the theory in such a pleasantly written 

 and well-published fashion, which will give it 

 in all probability a wide circulation. The 

 present review need comment on but a few of 

 its teachings — those which touch fundamental 

 conceptions in morphology. 



I. Gaskell maintains the doctrine that evolu- 

 tion proceeds (genetically) from the domi- 

 nant type of one geological horizon to the 

 dominant type of the following geological 

 horizon. This is a doctrine which at the best 

 is intangible and unconvincing. Even the ex- 

 ample cited by Gaskell does not support his 

 case — that the vertebrates in their earliest 

 occurrence superseded the sea-scorpions of the 

 early Paleozoic. For at that critical time it 

 was the oephalopods (e. g., the huge Ortho- 

 ceratids), not the sea-scorpions, which were 

 the dominant race. However, for the sake of 

 argument, granting that the cephalopoda are 

 lower than arachnids, they are obviously much 

 higher (measured by the standard of the 

 nervous system) than the lower crustaceans 

 or the worms, hence by the doctrine of domi- 

 nant types they should have taken an inter- 

 mediate genealogical position between the 

 Crustacea and the arachnids, which even Gas- 

 kell would deny. 



II. Gaskell seems to have little conception 



of parallelism, and he is probably, therefore, 

 unaware of the mighty literature dealing with 

 his theme. This is the more regrettable, since 

 it is this principle which bears so directly upon 

 phylogenetic studies. For it can now be dem- 

 onstrated beyond peradventure that animals, 

 e. g., of different orders, may develop similar 

 structures to such a degree that they are 

 sometimes mistaken for members of the same 

 family or even genus. And if this be true, 

 how can we believe that certain specified re- 

 semblances of king-crab to vertebrate can be 

 accepted as tests of genetic kinship? If 

 such a form as a Litoptern can develop many 

 essentials of a horse, yet be not included within 

 the great group of Ungulates, how can we 

 accept Gaskell's elaborate details when he 

 compares forms as widely apart as a vertebrate 

 and an arachnid? An arachnid has no tubu- 

 lar nervous system, no gill slits, no notochord ; 

 it has widely different appendages, skeleton, 

 skin, urogenital system, sense organs — how 

 therefore do we venture to compare its 

 details with such vertebrate structures as the 

 abducens, or the various branches of the fifth 

 nerve, or the jaw muscles? We might, in fact, 

 make a comparison of this kind as convinc- 

 ingly, or unconvincingly, by selecting the 

 structures of a cephalopod. In a word, it is 

 this kind of comparison which makes the dis- 

 tressed reader cast down Gaskell's book even 

 when assured by the author that the evidence 

 of these " homologies amounts almost to a 

 certainty." 



III. Gaskell fails to take into account a 

 fundamental rule in descent study, that series 

 of forms whose structures are the most closely 

 connected should be used for comparisons. 

 This rule he violates constantly when he com- 

 pares the Paleozoic Cephalaspis and the re- 

 cent lamprey-larva, Ammoccstes, for these 

 forms are by no means the nearest links in a 

 possible chain. As a matter of fact, we now 

 known that Cephalaspis is but one example 

 of a great group of Paleozoic creatures, 

 that these creatures show the greatest range 

 in forms and structures, and that many 

 of them are unlike arachnids even in super- 

 ficial regard, in fact some of them are soft- 

 bodied and covered with shagreen like a 



