14 GENERA OF THE SUBORDERS ORTHOIDEA AND PENTAMEROIDEA 



genera, Elefhas and Mastodon, and that not so long ago all vertebrate paleontologists regarded the 

 Proboscldea as a monophyletic line. Osborn, according to his "phylogenetic classification," now 

 divides the "phylum" Proboscidea into at least 1 6 races or lines of descent. These he arranges into 

 4 orders, 5 families, 13 subfamilies, 28 genera, and 290 species. 



A. Morley Davies in his preface to Sahni's very careful study of the terebratulids of the British 

 Chalk,^ a work wrought out along the best modern lines, has the following to say regarding the most 

 reliable generic characters in brachiopods: 



Internal structure has proved the surest clue to relationship. Beecher and others based the classification of 

 the "long-looped" Terebratellids primarily on the form and development of the brachial loop; Buckman relied on 

 the muscle-scars in the case of the Jurassic Terebratulids and RhynchoneUids . . . Once the definite clue has been 

 found by which distinct stocks can be separated, it becomes possible to discriminate among the multitude of seem- 

 ingly random external variations. Some of them, either individually or, more frequently, in particular associations, 

 are definitely correlated with internal features and become diagnostic for particular stocks: they belong to what 

 Douville has termed . . . "static characters," persisting or slowly changing as evolution proceeds. Others are 

 among Douville's . . . "progressive characters," which go through a series of stages often parallel in different stocks. 

 It is these parallel developments, sometimes predominating over the static characters so as to eflrect a "convergence" 

 or homoeomorphy between different stocks that are most apt to mislead in hasty determination. Yet these progres- 

 sive characters may sometimes remain relatively static for a period, so as to serve provisionally or partially as 

 diagnostic characters (pp. v-vi). 



When Sahni began his work there were 14 Upper Cretaceous species usually referred to Tere- 

 bratula. It now appears that not one of them belongs to this genus, his memoir defining 57 forms 

 in 12 new genera! This shows what must be done nowadays if one's work is to stand. One must 

 lift the covers and delve into the insides for the most reliable characters, and then carefully 

 examine the covers to discover the homoeomorphs and the significance of the slight changes in external 



form. 



Among students of brachiopods, none has more clearly set forth the best present methods of 

 determining the genus from the genetic standpoint than J. Allan Thomson.* He says: 



Biological classification is now an attemft to exfress the degrees of affinity through descent [italics ours]. 

 Instead of as few characters as possible being used to decide upon the generic position of a species, as many characters 

 as are necessary are employed to determine the relationship to other species. It has frequently been found that the 

 species grouped together under the older wide genera form homoeomorphous series of polyphyletic origin — u e., 

 their similarity of form has been attained through different lines of dissimilar ancestors (p. 120). 



According to Buckman (1918),® 



Generic division will not be complete for scientific purposes until all polyphyletic series are rightly separated. 

 If the criteria now obtained are insufficient for this work of separation, it is the task for future observers to note 

 and to utilize other characters (pp. 133-134). 



Thomson says, further: 



Another change in the conception of the genus has resulted from a fixation of procedure in regard to the 

 law of priority ... It has become necessary to define each group not only by an assemblage of characters 

 originally proposed for it by its author, but by a type within the group with which the aflBnity of other members 

 may be compared — a species by a type specimen or holotype, a genus by a type species or genotype, a family by a 

 type genus (p. 121). 



The genetic work of paleontologists must be guided, however, by the results of the neontolo- 

 gists, who have the living organisms to study in their entirety, but even among them the demonstra- 

 tion of lines of descent lies with the experimentalist. The work of these biologists shows how 

 deceptive "blood-relationship" may actually be. In this connection we may summarize here the con- 



' M. R. Sahni, The Terebratulids of the British Chalk, Men. Paleontog. See., 1927, 1929. 



* Brachiopod Morphology and Genera (Recent and Tertiary). N. Zealand Bd. Sci. and Art, Manual No. 7, 1927. 

 ° The Brachiopoda of the Namyau Beds, Northern Shan States, Burma. Geol. Surv. India, Pal. Indica, new ser., 

 vol. 3, mem. 2, 1917 (1918). 



