scHucHERT.] PRINCIPLES OF CLASSIFICATION. 115 



excluded, while the Ancylopoda contained all other brachiopods, both 

 articulate and inarticulate forms. These subclasses are further 

 divided, on the basis of the brachia, into four orders : Ancylobrachia, 

 Cryptobrachia, Sclerobrachia, and Sarcicobrachia. Of these the first 

 only has value as a superfamily, since it includes the "loop-bearing" 

 genera, or Terebratulacea. The other orders have so heterogeneous an 

 assemblage of forms as to be of no permanent value. 



Beyond the introduction of new fauiilies, no further attempt was 

 made by writers to divide the Brachiopoda into other orders than 

 Lyopomata and Arthropomata until 1883, when Waagen published his 

 great work on the fossils of this class from the Salt Rauge group of 

 India. He found it "absolutely necessary" to further divide the 

 Lyopomata and Arthropomata into seven suborders. The basis for 

 these suborders has no underlying principle of general application, yet 

 the majority of the divisions are of permanent value, for each contains 

 an assemblage of characters not to be found in any of the others. 

 Waagen's genealogy of the Arthropomata, with Orthis as the proto- 

 type, falls at once to the ground, since the comprehensive studies of 

 the genus Orthis by Hall and Clarke have shown that it is questionable 

 "whether any of these primordial forms can be included under Orthis 

 according to the strict definition of the term or even under any of the 

 subdivisions"' proposed by them. There are, however, a few species 

 in the Upper Cambrian which seem to agree with such dalmanellas as 

 0. sub(eqiiataj but these originated long after many undoubted Pro- 

 tremata and Telotremata had lived in the Lower and Middle Cambrian. 

 Lingula, on the other hand, was usually regarded as the prototype of 

 all brachiopods, but this is also imjiossible, since a number of inarticu- 

 late genera flourished for ages before Lingula was developed. 



PRINCIPLES OF CLASSIFICATION. 



No classification can be natural and permanent unless based on the 

 history of the class (chronogenesis) and the ontogeny of the individ- 

 ual. However, as long as the structure of the early Paleozoic genera 

 of Brachiopoda remained practically unknown and the ontogeny 

 untouched, nothing of a permanent nature could be attempted. In the 

 recent volumes by Hall and Clarke many of these early genera are 

 clearly defined, so that their structures and geologic sequence are now 

 far more accurately known. The ontogenetic study of Paleozoic species 

 was initiated in 1891 by Beecher and Clarke, and was continued by 

 Beecher and Schuchert. These results, combined with those derived 

 from the development of some recent species, and published by Kova- 

 levsky, Morse, Shipley, Brooks, Beecher, and others, confirm the con- 

 clusions reached through chronogenesis. Moreover, the application by 

 Beecher of the law of morphogenesis, as defined by Hyatt, and the 



1 Palaeontology of New York, Vol. VHI, Part I, 1892, p. 218. 



