the Pennsylvanian, and perhaps the late Pennsylvanian at 

 that. 



The problem of the origin of the limb does not seem to 

 be resolvable in terms of periodic drought. It is not likely 

 that a limb would be developed for periodic escape in 

 a population many of whose members inhabited permanent 

 waters. The fact that permanent waters existed, and on a 

 widespread scale, is indicated by the occurrence and wide 

 distribution of crossopterygians and early actinopterygians. 



The origin of the limb must be sought in some other 

 direction. It is more likely that amphibians developed in an 

 aquatic habitat where the limb was advantageous, not for 

 short periods but throughout the entire life of the organism. 

 The nature of the waters in which the amphibian devel- 

 oped would be different from that in which crossopterygians 

 and dipnoans lived. This is necessarily the case simply on 

 the basis of the observation that limbs did not develop in 

 these groups or that, at least, these groups remained fishes. 



It might be assumed that the limb developed as a means 

 for progression through vegetation-choked waters. Such 

 waters around the margins of ponds or along stream courses 

 were areas into which food organisms could escape from the 

 open-water predators. Such food reservoirs would be at- 

 tractive to an organism which could crawl through, around, 

 over, and under various obstacles. At this time, aquatic and 

 terrestrial vegetation was beginning to develop; certainly 

 vegetation-choked waters were common during the carbonif- 

 erous. Such waters, because of rotting, would require an 

 air-breathing type of animal. Invasion of this kind of habitat 

 could explain the development of limbs and air breathing 

 structures much better than the theory of movement over 

 dry land. The limb, as developed for the invasion of such 

 waters, would also serve for escape during periods of drying. 

 It was perhaps a more effective or preadapted device for pro- 

 gression over the land than the fin. 



The exact ancestry of the amphibian is open to question. 

 The general view is that an osteolepid crossopterygian was 

 the ancestral type since there is best agreement in terms of 

 the cranial roof However, as is evident from our review, 

 this supposition is not particularly well founded. There 

 appear to be striking resemblances between many features 

 of the amphibians and the dipnoans. These features are, 

 perhaps, just as important as the agreement in cranial 

 plates. The differences in the snout region between the am- 

 phibian and any choanate fish would suggest that this is a 

 separate line of evolution, which arose at the same time as 

 the other choanate lines. The common ancestor of all these 

 choanate organisms might have been somewhat closer to 

 the osteolepiform than to the dipnoan in some features, but 

 it would seem to be somewhere in between these two ex- 

 tremes in an over-all evaluation. 



As far as specific details are concerned, one might say 

 that the palatoquadrate of the ancestral tetrapod was closely 

 articulated with the endocranium, that the hyomandibula 

 was greatly reduced, and the pharyngohyal portion well 

 developed. The symplectic disappeared along with the re- 



duction of the hyomandibula, and the ceratohyal became 

 attached primarily to the palatoquadrate area. In this way 

 the many intermediate features of the hypothetical am- 

 phibian ancestor could be described in detail, but this seems 

 unnecessary. 



The view, then, that the amphibian is an osteolepid or 

 "crossopterygian" must be rejected on the grounds that, as 

 defined, the amphibian and crossopterygian represent paral- 

 lel, or largely contemporaneous, developments. The general 

 supposition that the crossopterygian, dipnoan, and am- 

 phibian all had a three-chambered heart at the time of 

 their separation, suggests that in some ways the amphibian 

 was the least modified of the three main groups and thus 

 was closer to the ancestral form. The crossopterygian, the 

 actinistian, and the dipnoan may represent modifications 

 away from the amphibian toward a more aquatic way of 

 life. 



These three examples are typical of the questions that 

 arise in systematics. In discussing them the nature of the 

 fossil record has been stressed, as well as the general role 

 of anatomical detail. Next, the place of embryology in a 

 phylogeny should be mentioned. 



THE BIOGENETIC LAW 



The so-called biogeneUc law, that ontogeny recapitulates 

 phylogeny, can now be reconsidered with more understand- 

 ing. As a generality, one can say that ontogeny supplies 

 little information on phylogeny. If we consider the early 

 stages of development as described in Chapter 7, it is ap- 

 parent that a general pattern of cleavage, blastulation, 

 gastrulation, and organogenesis occurs in each of the verte- 

 brates but each of these stages in this sequence of events 

 varies in each of the types described. The developmental 

 stages appear to be the most direct way by which a unicellular 

 egg can reach a multicellular, complex organisms. The 

 similarities of developmental stages reflect the similarities 

 of starting point and end product. 



From the many styles of development observed, it would 

 be difficult to say which is the more primitive and which 

 the more advanced. The development of the lamprey, for 

 example, is probably primitive in the form of the egg, its 

 cleavage, and gastrulation. The development of the nerve 

 tube from a solid neural keel may be either primitive or 

 advanced; the latter view is generally held. A similar type 

 of neural keel is observed in the actinopterygian fishes and 

 probably represents a parallel modification. It can be as- 

 sumed that an egg with a small amount of yolk is primitive, 

 and therefore that the lamprey is, perhaps, the most primi- 

 tive of the vertebrates in this respect; but in contrast to the 

 lamprey, the hagfish represents a highly modified form. 

 Thus, within the agnath fishes one sees both extremes, not 

 only in terms of the egg but also in terms of the neural tube. 



The variations of development as described indicate rather 

 clearly that each kind of vertebrate has modified different 

 stages of the developmental sequence to fit its own in- 



THE BIOGENETIC LAW • 445 



