L] ORIGIN OF HEREDITARY CHARACTERS 291 



How then are we to regard such hereditary features as those of the soral 

 slide, and the vascular fluting, decentralisation, and disintegration, from the 

 point of view of Descent? All of them suggest that features of adjustment 

 imposed upon a succession of individual lives by conditions, sometimes 

 external sometimes internal, have become hereditary. Naturally the reply 

 may be made that probably mutations favourable to the perpetuation of the 

 imposed character may have made that character permanent. If we grant 

 that, do we not thereby simply admit that the distinction between fluctuating 

 variations and mutations is not absolute? In other words, that fluctuating 

 variations repeatedly imposed adaptively upon successive generations, with- 

 out any near limit of time, are liable to become mutations? It is difficult to 

 see any other rational explanation of the wide-reaching facts of homoplastic 

 adaptation, shown in exceptional profusion in the ancient Class of the Fili- 

 cales, but evident also in Plants at large. 



This conclusion is, however, directly opposed to the opinion of those who 

 hold that characters impressed upon the individual life are not heritable. 

 A very fair and philosophical balance of their position as against the Mnemic 

 Theory of Semon, with a definite conclusion in favour of the latter, is to be 

 found in the Address of Sir Francis Darwin, as President of the British 

 Association in 1908. The Mnemic Theory proceeds on the conception that, 

 as a consequence of stimulus imposed by conditions of life, a record or 

 engram is impressed on the organism. Sometimes the engram may be 

 recognised only functionally: but Sir Francis remarks in further explana- 

 tion: "As I have attempted to show, morphological changes are reactions to 

 stimulation of the same kind as these temporary changes. It is indeed from 

 the morphological reactions of living things that the most striking cases of 

 habit are to be found." The examples given above from the Filicales appear 

 to be illustrations of this. 



Semon assumed that when a new character appears in the body of an 

 organism a new engram is added to the nuclei of the part affected; and that 

 further disturbance tends to spread to all the nuclei of the body, including 

 those of the germ-cells ; and so to produce in them the same change. But 

 this can only be made efficient by prolonged action. He laid great stress 

 upon the slowness of the process of building up efficient engrams in the 

 germ-cells. Direct experiments bearing on the inheritance of acquired or 

 impressed characters have so far given indecisive or negative results. But 

 all of these experiments range within the narrow limits of laboratory time. 

 If, however, reference be made to a sequence of events conducted with the 

 latitude of geological time, and the effect appears to be positive, it would 

 seem right to give such positive conclusions precedence over the negative 

 evidence of experiments limited within a brief period. This is the interest 



