1910.] Gill's Principles of Classification. 91 



GILL, 1870, 1872. 



'On the Relations of the Orders of Mammals.' Proc. Amer. Assoc. 

 Adv. Sci., 1870, 19th meeting, pp. 267-270. 

 'Arrangement of the Families of Mammals,' Smithsonian Miscellaneous 



Collections, 1872. 



The early classifications of the Darwinian epoch revealed an extreme 

 reliance on single characters which Linnaeus, Cuvier and de Blainville had 

 wisely avoided. Thus Haeckel, as we have seen, developed the most elabo- 

 rate phylogenetic classifications on the primary basis of jilacentation, while 

 Owen, going far beyond de Blainville, had selected brain characters as funda- 

 mental, and upon that assumption had erected three "subclasses" within the 

 limits of the Placentalia. In view of these considerations it is interesting to 

 find the present "Dean of American Taxonomy" turning in 1870 to a more 

 normal development of Linniiean methods, and producing an arrangement 

 of the orders which is remarkable for its simplicity, its selection of the best 

 features of preceding classifications and for the lucid statement of guiding 

 principles. The latter are, in fact, so illuminative, that it may be permitted 

 to quote them in full. 



"1st, Morphology is the only safe guide to the natural classification of 

 organized beings; teleology, or physiological adaptation, the most unsafe 

 and conducing to the most unnatural approximations. 



"2d, The affinities of such organisms are only determinable by the sum 

 of their agreements in morphological characteristics, and not by the modi- 

 fications of any single organ. 



"3d, The animals and plants of the present epoch are the derivatives, 

 with modification of antecedent forms to an unlimited extent. 



"4th, An arrangement of organized beings in any single series is, there- 

 fore, impossible; and the system of sequences adopted by genealogists may 

 be applied to the sequence of the groups of natural objects. 



"5th, In the appreciations of the value of groups, the founder of modern 

 taxonomy (Linnaeus) must be followed, subject to such deviations as our 

 increased knowledge of structure necessitates. 



"The adoption of such principles compels us to reject such systems as are 

 based solely on modifications of the brain, those of the placenta, and those 

 of the organs of progression, such modifications not being coincident with 

 corresponding modifications of other organs, and therefore not the expres- 

 sions of the sum of agreements in structure." 



Some of the- more noteworthy features of the classification are as follows: 

 (1) The return to de Blainville's three grand divisions; (2) the grouping of 



