245 



have to be careful, or the ease and abundance with which 

 the rogue forms assert themselves, and are thus liable to be 

 put forward for sale, may water down the popular interest 

 now felt and revive to some extent the old idea that British 

 Ferns are not worth growing. 



C. T. D. 



FERNS AND MENDELISM. 



I have sometimes been asked whether I considered that 

 Ferns were subject to Mendelian laws of heredity, but 

 although I have no doubt that those laws pervade all 

 organic life and embrace therefore Ferns, it appears to me 

 a matter of extreme or indeed insuperable difficulty to 

 establish this as a fact in the same way as has been so 

 abundantly done in the animal and plant world generally. 

 The reproductive operations of the Fern appear to me to 

 be conducted on such lines as practically to exclude 

 scientific certainty in tabulating results or even in tracing 

 results to definite causes. The well-established fact that 

 when once a fern has " sported " its progeny are liable to 

 sport again, introduces a disturbing factor, as indeed the 

 faculty of spontaneous " sporting " must do in all Mende- 

 lian experiments, by introducing new " characters " which 

 may entirely upset the calculations of the experimentalists. 

 Another grave difficulty is involved in the minute 

 character of the reproductive operations, which renders 

 crossing on systematic lines a practical impossibility. The 

 only line to be followed by would-be investigators is to 

 ignore varieties and to confine their efforts to the crossing of 

 normal forms of species sufficiently closely allied but also 

 sufficiently distinct in character to permit of recognition of 

 such distinctive characters in the crossed progeny should any 

 result. Even here, however, should F^, or the first family, 

 produce recognisable results, which it might not, even with 

 a successful cross, since the dominant character usually 

 asserts itself in Fi, the cross, thus showing that type only, 



