326 THE VARIATION OF ANIMALS IN NATURE 



innate tendencies. In so far as the latter are not demonstrable 

 except by their results, this use of Orthogenesis is admittedly 

 an appeal to ignorance. But an appeal to an unknown activity 

 (which after all is by no means absent from other theories of 

 evolution, nor indeed from any theorising on vital activities) 

 is not necessarily inadmissible, especially if the other available 

 explanations are ruled out or shown to be implausible. 



The bulk of the writers who have espoused the orthogenetic 

 standpoint have perhaps wisely but timidly confined themselves 

 to the description of facts. Eimer's ' Laws of Organic Growth ' 

 (organophysis), for example, are actually merely generalised 

 from observation and are not in any sense a causal theory. 

 Besides such writers as have sought a general explanation in the 

 pressure of the environment, Dendy (191 1), Champy (1924), 

 and Lang (1921) have faced the necessity of supplying a causal 

 explanation of the particular orthogenetic phenomena they 

 studied. Fenton {I.e.) has attempted to harmonise the parti- 

 cular phenomena of recapitulation with theories of the indi- 

 vidual life-cycle put forward by Child and others. 



The facts of parallel variation enter into this discussion 

 rather at second hand and are not directly relevant to the 

 question as to whether a determinate evolution, undirected 

 by selection, occurs or not. They are relevant to this extent, 

 however, that if they are not attributable to similar selective 

 agencies or similar environmental stresses, their occurrence is 

 an indication of the limitation of the evolutionary potentialities 

 of animals. In plants parallel variation is common enough 

 to form the basis of Vavilov's law of ' Homologous Series.' 

 In animals instances are to be found in Eimer's own work on 

 Lepidoptera (1897), in Gadow's observations on the pattern 

 of Coral Snakes (191 1), and in Parker's study of Brevicipitid 

 Frogs (1932). Such series certainly necessitate a modifica- 

 tion of the conception of an all-round variability, but the 

 mere fact of their occurrence does not necessarily involve 

 the conclusion that they are non-adaptive. That conclusion 

 could be arrived at only by an examination of the value of 

 the characters on their own merits. Annandale and Hora 

 (1922) and Prashad (1931) have clearly shown that parallel 

 evolution of adaptive structures occurs in exceptional habitats. 



It will be seen that we have three classes of phenomena 

 that have been treated as ' orthogenetic ' on the grounds 



