HOMOLOGY, ANALOGY AND PLASIS 



171 



fort spent on the details of neurological 

 anatomy is justified only when the work 

 is meant to be brought into correlation 

 with the physiological. What holds 

 true of a pure morphology of the brain 

 holds equally true of morphology gen- 

 erally. Morphology finds a meaning only 

 in its functional setting. At our present 

 stage of history it is easy for us to see that 

 the original decision to separate physi- 

 ology from comparative anatomy was 

 foredoomed to failure. 



In view of the preceding exposition it 

 is plain that the broad process here called 

 plasis, the process to whose investigation 

 the experimental embryologists have al- 

 ready devoted themselves in the field of 

 ontogeny, is equally insistent in the realm 

 of phylogeny. We have seen how Lan- 

 kester, a Darwinian and believer in natural 

 selection, was early constrained to postu- 

 late special environmental "forces" in 

 order to account for certain aspects of 

 phylogenetic plasis. There can be little 

 doubt that plasis is a controlled process 

 and proceeds according to law. Granted 

 the fact of evolution, the astounding fit 

 between the animal as a whole and its 

 environment, and the correspondingly 

 wonderful internal adjustment of all 

 its diverse parts one with another, seem 

 to indicate the operation of (physical) 

 law. Endlessly complex as animal struc- 

 ture is, the dominant principles involved 

 in maintaining the structure or in chang- 

 ing it according to need, may be few and 

 relatively simple. 



Roughly speaking and without attempt- 

 ing at the moment any further sifting of 

 detail, we may recognize three (or four) 

 elements in phylogenetic plasis: (1) the 

 power to transmit by heredity the tempor- 

 arily existing condition of structure; (2.) 

 adaptive modification of existing struc- 



ture; (3) the invention of new structures; 

 to which we may add, whether an 

 independent thing or not, (4) the occa- 

 sional dropping of existing characters. 

 Thanks to the work of the school of 

 T. H. Morgan and to the investigations of 

 Goldschmidt, our knowledge of the 

 mechanism of (1) is rapidly becoming 

 more precise and definite. We may also 

 have some possible knowledge of the 

 mechanism of (4), but scarcely yet of (4) 

 in its adaptive implications. The condi- 

 tions that control (2.) and (3) remain 

 quite obscure, and these are the very 

 elements that have not yet been subjected 

 to the rigorous physiological mode of 

 inquiry. Lucas has sought to sketch a 

 method whereby one might begin with the 

 investigation of (3). In the course of our 

 present exposition we have obtained a 

 physiological test for homology, which 

 implies that we have a better means 

 than before of observing the operation 

 of (x). 



The failure up to the present to establish 

 by experiment the transmission of "ac- 

 quired" characters may well be due to the 

 paucity of our knowledge respecting (2.) 

 and (3). It has often been suggested that 

 our laboratory attempts to cause phylo- 

 genetic modification may not have taken 

 account of an adequate time factor. 

 Apart from the question of time there is 

 another possible way of looking at the 

 matter. Any given animal is a system in 

 equilibrium with its environment, and is 

 built up of a vast congeries of systems all 

 in equilibrium with each other. To use 

 a very rough illustration, the organism 

 might conceivably be likened to a complex 

 jointed linkage, which is poised or 

 balanced at some central region, all its 

 various outlying parts being similarly 

 poised but so interconnected that an 

 adequate disturbance of one or more may 

 cause a rearrangement of the whole with 



