Inheritance of Acquired Characters. 97 



weight of the roof ; so alternate breakdown and repair have led to 

 the present structural relationships in many parts of the human 

 body. Thus the heart is described as an aneurism, and the stomach 

 as a dilatation sac, whose thick muscular wall has developed as the 

 result of disaster and repair. It is assumed that the transmission 

 of these changes probably takes place by a morphogenetic reply 

 in utero to increased functional stresses. 



Brief mention must be made of the important work of 

 Harrison (1920a) on melanism and other racial characters in 

 certain Geometrid moths. After intensive observation and 

 experiment with the genus Oporabia, Harrison concludes that the 

 subspecies O. filigrammaria was evolved from O. autumnata during 

 the Glacial period by the action of changed climatic conditions, 

 and that " almost certainly" many of the racial and subspecific 

 differences are " true Lamarckian effects," in particular the food 

 instincts in 0. filigrammaria and the period of emergence of the 

 pinewood race of O. autumnata. Moreover, a Lamarckian 

 explanation of melanism is considered necessary, the darkening 

 resulting from metallic salts in the food, derived from the smoke 

 in industrial areas and from sea fogs in coastal areas where 

 melanism frequently occurs. Melanism is known to behave as a 

 simple Mendelian dominant in many species (see e.g., Onslow, 1920). 

 Harrison (1920b) has recently shown this to be true of the melanic 

 variety of Tephrosta crepuscularia when crossed with the type, out 

 in interspecific crosses between T. crepuscularia and T. bistortata a 

 chaotic series of F 2 and F, forms was obtained, the behaviour 

 being no longer describable in Mendelian terms. This recalls the 

 result obtained by the writer (1915f) in crosses between the 

 Mendelian dominant character in (Enothera rubricalyx and another 

 species, CE. grandiflora. In both cases dilution and modification of 

 the character has resulted from crossing with a different species. 

 If crossing will modify such a unit-factor, then it is not unreasonable 

 to suppose that they may be capable of environmental modification. 

 If the Lamarckian explanation of melanism is a true one, then 

 it would appear to be a case of parallel induction, because of its 

 manner of inheritance in crosses, indicating that the nuclear 

 structure is already altered. The function of hormones as an 

 evolutionary means of handing on or accelerating modifications, as 

 suggested by Cunningham (1908), Dendy (191 1) and others, indicates 

 a method by which modifications may be perpetuated until the 

 change becomes germinal by producing an alteration in the nucleus. 



