APOGAMOUS FERNS. EVOLUTION OF THE SEPARATE SPECIES 

 an origin by simple chromosome doubling. Pairing in the diploid, on the other hand, is 

 virtually absent (see Fig. 200 a). In certain cases (e.g. var. polydactyla Dadds) a triploid 

 may also show complete failure of pairing, but in the more normal forms of the species 

 some pairs are present in triploids, suggesting that not all the triploids found wild are of 

 the same genetical nature; some may perhaps be primitive triploids (i.e. the ancestral 

 type of the species whatever that may have been), while others (e.g. polydactyla Dadds) 

 may be triploids of secondary origin such as could be formed by a cross between an 

 apogamous diploid and a related sexual diploid. Until more is known about the nature 



% 



:^' 



c d 



Fig. 20 1 . Spores of the polyploid series in Dryopteris Borreri Newm. from glycerine jelly mounts. 



a, diploid; b, triploid; c, tetraploid; d, pentaploid. 



X 1000. 



of related sexual species this problem cannot be solved, and it is equally impossible to 

 state with assurance which of the two first members of the series (diploid or triploid) 

 is in fact the older, for while a derivation of the latter from the former can have occurred 

 along the fines just indicated, a reverse derivation is also conceivable. Should any of 

 the spores produced by the sixteen-celled sporangia be viable, we know enough about 

 the breeding behaviour of triploids (cf. Osmunda, Chapter 3) to be certain that the 

 viable types could not fail to include those in which the regular diploid chromosome 

 number had by chance been reassembled.* 



That the production of an occasional viable spore from a sixteen-celled sporangium 



* Since the above was written some positive evidence in favour of the primitive form of D. Borreri 

 having been diploid has come to hand by finding that the population of this species on the unglaciated 

 island of Madeira is exclusively diploid. 



192 



