398 W. Bateson. 



type would be seen to return with its appropriate frequency. The 

 further question then suggests itself whether the cases in which a 

 parental type has not hitherto been observed to re-appear in F«, may 

 not in reality be extreme cases of the same phenomenon. 



In the beetles lAna Lapponica, and Gastroidea dissimilis Miss 

 Mc C r a c k e n (74) has described several peculiarities. In the former 

 species the spotted type is dominant to a black form, and in the 

 latter, metallic green is recessive to metallic black. Several peculiar 

 phenomena of transmission are said to occur in these species, the 

 details of which must be studied in the original papers. The tables 

 given suggest a doubt whether the analysis was complete in all 

 particulars and it is to be hoped that a fuller study of these re- 

 markable cases will be made. 



In Antirrhinum de Vries (Mutationstheorie, II, p. 76) states that 

 dwarf-forms extracted from crosses with tall do not breed true. If 

 confirmed, this fact will be of much interest. 



Perhaps the most curious of the unconformable cases yet certainly 

 established is that which concerns the inheritance of double flowers 

 in MatthioJa. In numbers of individuals doubleness, as studied by 

 Miss Saunders has been found to be an ordinary recessive, as it 

 is also in Primida Sinensis}) Nevertheless it is well known that 

 double Stocks, being totally sterile, without pollen or seeds, are always 

 produced from singles. The proportion of doubles given by such plants 

 is sometimes very high, being often 60, and occasionally even 80 per 

 cent. How it happens that a recessive type can be produced in these 

 high proportions is as yet quite unknown. 



The intercrossings carried out by de Yries among the wonderful 

 mutational forms raised by him from Oenothera have given rise in 

 many cases to results which are undoubtedly very anomalous, but until 

 these cases have been analysed more fully and studied in the light 

 of recent knowledge of the inter-relations which may subsist between 

 characters, it is difficult to be sure as to the significance of these 

 phenomena. A better knowledge of these curious cases would pro- 

 pably lead to important advances in genetics. 



First-crosses breeding true. 



Of the considerable number of examples in which this phenomenon 

 has been alleged to occur, few have been satisfactorily observed on a 

 large scale. A discussion of some of these cases is given by de Vries 

 in Mutationstheorie (especially II, p. 66). The evidence is not in my 



*) From evidence given by Mr. Douglas to the Hybridisation Conference, 

 1906 it appears likely that in the Carnation doubleness is a dominant character. 



