478 BOTANY OF THE LIVING PLANT 



as " somatic parthenogenesis," which impUes that the embryo springs from the 

 ovum, but the ovum was itself diploid. A like condition has been seen in 

 Marsilia Drummondi, and in Athyrium filix-foemina, var. clarissima. In 

 such cases again no fertiHsation is necessary to arrive at the diploid state. 

 A more rare condition is that where an egg that is really haploid develops 

 as though it had been fertihsed. This is not known in Seed-Plants, but it 

 has been found to occur in Chara crinita. It is described as "generative 

 parthenogenesis." 



Contrasted with these cases of apomixis, where no sexual fusion occurs, 

 are various conditions which may be ranked as substituted sexuality. Here a 

 nuclear fusion is seen, with consequences like those following on normal 

 syngamy. But the nuclei involved are not produced in the normal way. 

 Examples have already been described for Nephr odium pseudo-mas, var. poly- 

 dactylum (p. 350, Fig. 294) ; and in that unusual type of nuclear association 

 which occurs in the formation of the fruit in the Uredineae (p. 448, Fig. 384). 

 In the latter the association is ultimately followed by nuclear fusion, 

 which thus, though deferred, takes its place in the cycle. The fact that 

 similar nuclear fusions precede the formation of ascospores (p. 430) and of 

 basidiospores (p. 441) suggests that such methods are probably wide-spread 

 among Fungi. The nuclei involved in such cases spring not from any widely 

 distinct sources, but from cells closely related in position and in origin. It is 

 rem.arkable that these and other irregularities in the sexual cycle are found 

 commonly among Plants represented by very numerous closely related forms. 

 If their numerous species and varieties have resulted from mutation, then it 

 would appear that excessive mutation may have had some influence in pro- 

 ducing those irregularities. In organisms which show mutation freely, the 

 INIendelian sifting out, and preservation of each heritable mutation, would be 

 a less vital matter than it is in more stable forms. This may in some degree 

 account for such deviations from the normal sexual reproduction as have just 

 been described. 



I 



