304 AN INTRODUCTION TO MODERN GENETICS 



Finally, we come to experiments especially undertaken to prove or 

 disprove the inheritance of acquired characters. The experiments are 

 too many to be fully discussed here. Very many of the experiments 

 must be dismissed for lack of adequate controls . It is essential to show 

 that any effects which may be produced cannot have been due to 

 selection, and this can best be done by using only genetically homo- 

 zygous stocks in which there is no basis for selection to act on. As an 

 example of an experiment which fails in this respect, we may take 

 work of Harrison^ on gallflies {Pontania salicis). These normally lay eggs 

 on the willow Salix Andersoniana and allied species. Harrison liberated 

 some in a district where the only available willow was S. rubra, and 

 found that after five years the flies had become acclimatized to this 

 species and refused S. Andersoniana when it was offered them. In the 

 first two years of the experiment, however, most of the galls laid on 

 *S. rubra aborted, and it is therefore likely that in the initial stages there 

 was stringent selection for individuals with a taste for this plant. It is 

 also possible that in this experiment the kind of food plant selected by 

 a female for oviposition is determined not by any direct genetic mechan- 

 isms but by the adult's memory of its own larval food. In some other 

 experiments the selection is even more obvious. For instance, Diirken^ 

 found that if pupae of Pieris brassicae are exposed to orange Ught some 

 of them become green. If these green pupae are bred from, a still 

 higher proportion becomes green, and the high proportion of greens 

 persists in later generations even without the orange Hght. This is 

 clearly an experiment in selection rather than in the inheritance of 

 acquired characters. 



There are two outstanding experiments which require more serious 

 consideration. Harrison and Garrett^ claimed that if certain moths are 

 fed on foliage impregnated with lead or manganese salts, such as occurs 

 in many industrial districts, melanic forms are produced by gene 

 mutation and breed true as ordinary recessives. This could provide an 

 explanation for the melanism so commonly found in moths in industrial 

 areas. It should be noted, however, that the phenomenon is not a 

 straightforward Lamarckian one, since there is no suggestion that the 

 melanism is adaptive. The efifea postulated is a direct chemical attack 

 on the germ plasm, producing specific mutations. Other attempts to 

 alter the genes by chemical means have usually been unsuccessful.* 

 Thermal and X-ray effects on the mutation rate are of course well 



^ Harrison 1927. ^ Diirken 1923. ' Harrison and Garrett 1926. 



* Sacharoff 1935. 



