CHAPTER IV 



FERN SELECTION 



Having in the last chapter given advice as to how to propagate 

 by buds, patent or latent, and by spores, we may now consider 

 what is the best material, especially in the latter connection, to be 

 used, in order to improve quality and enhance the charm, in which 

 direction so much has been done in the past, and so much may be 

 done in future. It is a curious fact that once a Fern, or indeed any 

 other plant, has departed from the beaten track, i.e. the normal or 

 common form, and as a " sport " has adopted a different habit of 

 growth, shape of foliage, or other practically new character, it is 

 apt to display these same characters in its progeny, i.e. breed fairly 

 true, and sometimes entirely true, throughout the brood. In many 

 instances, however, careful examination of the seedlings or spore- 

 lings will show divergences, some in the direction of the normal 

 form, i.e. reversion, though very rarely entirely, and others in a 

 forward direction, showing the " sport" character more markedly, 

 and it is, of course, in this direction that the judicious spore sower 

 should steer. Given improved seedlings (we prefer seedling to 

 sporeling, even in Ferns, since practically a seed precedes the young 

 Fern) of this class, the probability is that their offspring will vary 

 still more, and we may, in this connection, point especially to the 

 immense number of beautiful Hartstongues which are now ex- 

 hibited at our chief floral shows, as examples which, if sown from, 

 would probably yield no two plants exactly alike, though all would 

 be on similar lines. In fact, the great majority of them exemplify 

 this in themselves. To propagate such truly we must resort to 

 division or the induced bulbils already described. Ferns which 

 present any signs of inconstancy, such as partial reversion, or any 

 irregularities of form, defective pinnae, etc., should be strenuously 

 avoided by the spore sower, since defects are almost invariably 

 transmitted, and there is nowadays such a wealth of unexception- 

 able material that it is waste of time to attempt reformation of 

 what are known in Fern language as " rogues." Thoroughbreds, 

 on the other hand, i.e. perfectly symmetrical and constant forms, 

 may be relied on, and in some instances, three especially, with 



27 



