460 



SCIENCE 



[N. S. Vol. XXXVIII. No. 



homologies date from Owen; Gegenbaur 

 and Haeckel rearranged them slightly. 

 Lankester, in 1870, introduced the terms 

 homogenous, meaning alike born, and 

 homoplastic or alike molded. Mivart 

 rightly found fault with the detailed defi- 

 nition and the subdivisions of homoplasy, 

 and very logically invented dozens of new 

 terms, few of which, if any, have survived. 

 It is not necessary to survey the ensuing 

 literature. For expressing the same phe- 

 nomenon we have now the choice between 

 homoplasy, homomorphy, isomorphy, het- 

 erophyletic convergence, parallelism, etc. 

 After various papers by Osborn, who has 

 gone very fully into these questions, and 

 Willey's "Parallelism," Abel, in his fas- 

 cinating "Grundziige der Palseobiologie, " 

 has striven to show by numerous examples 

 that the resemblances or "adaptive forma- 

 tions" are cases of parallelism if they de- 

 pend upon the same function of homologous 

 organs, and convergences if brought about 

 by the same function of non-homologous 

 organs. 



I suggest an elastic terminology for the 

 various resemblances indicative of the de- 

 gree of homology of the respective organs, 

 the degree of affinity of their owners, and 

 lastly the degree of the structural likeness 

 attained. 



Homogeny. — The structural feature is 

 invented once and is transmitted, without 

 a break, to the descendants, in which it 

 remains unaltered, or it changes by muta- 

 tion or by divergence, neither of which 

 changes , can bring the ultimate results 

 nearer to' each other. Nor can their owners 

 become more like each other since the re- 

 spective character made its first appearance 

 either in one individual, or, more probably, 

 in many of one and the same homogeneous 

 community. 



Homoplasy. — The feature or character is 

 invented more than once, and indepen- 



dently. This phenomenon excludes abso- 

 lute identity; it implies some unlikeness 

 due to some difference in the material, and 

 there is further the chance of the two or 

 more inventions, and therefore also of their 

 owners, becoming more like each other than 

 they were before. 



CATEGORIES OP HOMOPLASY 



Isotely. — If the character, feature or 

 organ has been evolved out of homologous 

 parts or material, as is most likely the case 

 in closely related groups, and if the sub- 

 sequent modifications proceed by similar 

 stages and means, there is a fair probability 

 or chance of very close resemblance. Iso- 

 tely: the same mark has been hit. 



Homaotely. — Although the feature has 

 been evolved from homologous parts or 

 material, the subsequent modifications may 

 proceed by different stages and means, and 

 the ultimate resemblance will be less close, 

 and deficient in detail. Such cases are most 

 likely to happen between groups of less 

 close affinity, whether separated by dis- 

 tance or by time. Homozo-tely : the same 

 end has been fairly well attained. The 

 target has been hit, but not the mark. 



Parately. — The feature has been evolved 

 from parts and material so different that 

 there is scarcely any or no relationship. 

 The resulting resemblance will at best be 

 more or less superficial ; sometimes a sham, 

 although appealing to our fancy. Para- 

 tely: the neighboring target has been hit. 



EXAMPLES 



Isotely: 



Bill of the ArdeidcB balwniceps (Africa) and 



Cancroma (tropical America). 

 Zjgodactyle foot of Cuckoos, Parrots, Wood- 



pecTcers (2.3/1-4). 

 Patterns and coloration of Elaps and other 



snakes. 

 Parachute of Petaurus (marsupial) ; Pteromys 



(rodent) and Galeopithecus. 

 Perissodactylism of Litopterna and Hippoids. 



