Mutations in Other Fauiilies. 



297 



One of two instances will serve to exemplify these 

 generalizations here. I shall reserve a more complete 

 proof of them until I come to the description of the 

 separate new species. 



Pure sowings of 0. Icptocarpa, O. nancUa and 0. oh- 

 longa gave rise solely to their own type with the excep- 

 tion of the following mutants. Subjoined is a list of 

 these, together with the date of the experiment and the 

 number of the seedlings. 



MUTANTS ARISING FROM NEW SPECIES. 



NUMBER OF 



That is to say a proportion of about 0.2 %, whereas O. 

 Lamar ckiana usually has 1-3 % and often more mutants. 



O. scintillans is an example of an inconstant new spe- 

 cies in as much as often only about Ys of its offspring 

 are scintillans (see p. 245) ; the two remaining thirds 

 are composed of partly 0. ohlonga and partly 0. La- 

 marckiana and, to a much less extent, of the two other 

 common species 0. lata and O. nanella. 



I found the following numbers of these in the experi- 

 ments with O. scintillans in which the seeds were set on 

 typical plants which had been fertilized by their own 

 pollen. 



