MUTATIONS AND EVOLUTION. 303 



limited as regards their somatic effect, which have arisen quite 

 apart from any adaptive considerations, and are therefore beyond 

 the reahn of natural selection, and have no bearing on the evolu- 

 tion of the ostrich. However small the changes may be, they are, 

 nevertheless, held to represent discrete factorial differences, and, 

 following the views of Prof. T. H. Morgan, must be comprised 

 under the category of mutations or saltations. Similar varia- 

 tions occur in both the northern and southern birds, so that 

 corresponding germinal changes must have recurred in different 

 members or strains of the race. In any region intermingling of 

 the different types has freely taken place, but, tmtil artificial 

 selection was introduced, no one line of variation presented any 

 particular ascendency over another. Recently domesticated birds 

 in East Africa, reared directly from wild chicks and uninfluenced 

 by any farming selection, have been found to be just as variable 

 in their plume characters as are those in North and South Africa. 



Racial of Specific Mutations. 



Whether we regard the North African and the South African 

 ostrich as two distinct species, or only as sub-species or varieties, 

 each has a combination of well-defined differences which are 

 germinal in origin. These variations are manifestly of a dif- 

 ferent category from those above mentioned, and indicate fac- 

 torial changes of a more embracive nature. The northern bird 

 is larger than the southern, and diff'erently coloured, and displays 

 a bald head in place of one wholly covered with hair-like 

 feathers ; the hen also lays larger and perfectly smooth eggs 

 compared with those of the Cape bird, which are strongly pitted 

 and more oval in shape.* The characters of the northern bird are 

 retained under southern conditions, and reappear in the progeny ; 

 also the two freely interbreed, and the hybrids are fertile in 

 every respect. The bald head patch behaves as a homozygous unit 

 character, being dominant in first crosses and segregating nor- 

 mally both in the F^ generation, and when crossed with the 

 southern recessive. In first crosses the other distinguishing 

 characters — dimensions, colours and egg-shell — appear as inter- 

 mediates of varying degrees, but the slow attainment of maturity 

 of the ostrich, at about three years, has not yet permitted of their 

 analysis in later crosses. 



With full knowledge of the habits of the ostrich and of the 

 many vicissitudes to which it is subject in nature, it is hardly 

 possible to conceive of any circumstances in which the characters 

 distinguishing the two species could be of any adaptive or utili- 

 tarian value. The wild bird is subject to many harmful in- 

 fluences from the egg stage onwards, such as carnivora, droughts 

 and parasites, with none of which the features mentioned could 

 have any concern. They appear to represent merely isolated 



* "Journal of Heredity," vol. ix, Oct., 1918, p. 243. 



