1902.] NATURAL SCIENCES OF PHILADELPHIA. 213 



nephridia, nerve coi'ds, gonads, gills, etc." Thex-e are finally parts 

 of these organs. 



Now we find from this brief examination that there are different 

 degrees of structui-al parts in the organization, beginning with the 

 general body plan and proceeding to the minor parts which compose 

 it. Shall the relationship of two organisms be determined by the 

 comparison of their body plans, or of their major organ complexes, 

 or of the smallest complexes ? And how is this comparison to be 

 carried out, and on what kind of parts is the comparison to be 

 instituted ? We may immediately proceed to the discussion of these 

 questions. 



(b) The Criterion of Morphological Value. 



It is not justifiable to give any a priori decision as to what parts 

 are of value in determining questions of relationship, since the 

 criterion must be empirically souglit. 



As essential similarity is interpreted as denoting genetic affinity, 

 so also there is a second principle in determining phylogeny, namely, 

 that conservatism or fixity of a structure implies morphological 

 value. By conservatism or fixity is implied the continuance of a 

 part with relatively little change through a long line of evolution. 

 If in a line of evolution leading from a type of organization ^ to a 

 type N, a certain structure should be found in N little changed from 

 its appearance in A, it would be called conservative; if it had 

 become much modified or even lost before the stage of type N, it 

 would not be conservative. The value of a structural characteristic 

 for phylogenetic purposes depends directly upon how fully it con- 

 serves ancestral characteristics, and the less completely it conserves 

 them the less value it has. This is the well-based working proce- 



u •< Organ " is employed in comparative anatomy in a very loose sense, 

 as is necessitated by the practical impossibility of sharply distinguishing 

 the intergrading kinds of structural elements. As generally used for the 

 Metazoa, it is not applied to an autimere or to a metamere, but to more 

 segregated portions of them. Physiologically, it is a part with a partic- 

 ular function to perform ; morphologically, a specialized part of partic- 

 ular structure. While no sharp definition "is possible to cover all cases, we 

 find that "organ" is generally employed by phylogenists in the sense 

 of "the largest specialized part in an organization performing a particu- 

 lar function"; thus the respiratory organ of a vertebrate would be not 

 a lung cell, but the whole complex of lung cells. And it is well to 

 define the various degrees of structural jiarts as closely as possible, in 

 order to compare in different animals only such parts as relatively 

 correspond. 



