290 



infertility, suggest the idea that the original find may be a 

 cross between P. angulave and P. acnieatnm, since some of 

 the plants which did not present the " gracillimum" 

 character bear undoubtedly the plainly-stalked pinnules of 

 P. angtilave instead of the wedge-shaped unstalked ones 

 of P. acnieatnm^ and also show the duller surface of 

 angulave. The majority of the seedlings, however, are true 

 to the parental (p til chew imum) form, so that the original 

 object of the sowing has also been attained plus the un- 

 expected and highly gratifying one. 



Curiously enough, seedlings raised from the reversion 

 just described for the purpose of testing whether the 

 " gracillimum " tendency would be transmitted have 

 yielded plants which, so far, all appear to be of aculeatum 

 type rather than angulave, which the immediate parent 

 resembled. They are not, however, as yet sufficiently 

 developed to permit of certainty in this respect. Outside 

 the three remarkable cases we have described there are, of 

 course, minor ones. Mr. Wilson's A. f.f. setigevum, for 

 instance, with its bristly pinnae, has — whether by crossing 

 or otherwise is not clear — contributed, at any rate, to form 

 a bristly section of crested forms of very great beauty, 

 though, in some cases, rather addicted to reversion. We 

 have one plant of fine percristate type which has split into 

 two distinct crowns, one of which bears percristate fronds 

 of precisely the same outline and make but utterly devoid 

 of a trace of " setigerum," a clear proof, to our mind, of 

 reversion to a plain-crested parent. A. f.f. Frizellia, also, 

 probably in the same way, gave rise to a large number of 

 crested varieties on diverse lines ; but they, too, are 

 notoriously given to reversion, and, there is little doubt, for 

 the same reason. Both these types would appear to cross 

 extra freely with others of the same species, the two con- 

 joined types being apt subsequently to separate, a sort of 

 reversion resulting. A marked example of this "gay 

 Lothario " tendency is seen in Colonel Jones' P. ang, foly- 

 dactyliim, which appears to cross with any variety it is 



