CONCLUSIONS 



species could be investigated and especially if breeding work could be carried out. 

 Here, if anywhere in the Pteridophyta, we might hope to learn something about the 

 mechanism involved in aneuploid changes. 



(7) Chromosome size was commented upon in Chapter 16, but the possible signi- 

 ficance oUhanges o{ ?>\x&, especially in association with advancing polyploidy, should per- 

 haps be listed here. It can frequently be observed {Ophioglossum is a good example) that 

 if related species are compared, the one with the higher number will often possess the 

 smaller individual chromosomes and the reverse has not been found. This suggests 

 that diminution of chromosome size must often either accompany or follow the inci- 

 dence of polyploidy, and since our experience with Osmunda gave no indication of the 

 former, it seems possible that the order of events is the latter. The interest of this obser- 

 vation is in its possible relation to the paradox enumerated under item (2) above, 

 namely, that the grades of polyploidy occurring naturally seem incommensurate with 

 the limits encountered in artificial series. It seems probable that some nuclear or 

 physiological readjustments must occur with the passage of time to restore the power of 

 an organism to sustain a repetition of polyploidy on a scale which would be impossible 

 otherwise. We know nothing of the nature of such readjustments, but diminution of 

 chromosome size is perhaps one. 



This enumeration of evolutionary mechanisms expresses the factual basis for the com- 

 parison of the Pteridophyta with the Cruciferae, though it does not wholly exhaust the 

 general conclusions which can be drawn, some of which will next be discussed. 



An observation which is strengthened by the facts in both groups is the difference in 

 evolutionary effect of aneuploid and polyploid changes. The latter make species only. 

 The former make species also in the first instance, but such species seem usually to be 

 potential genera or larger groups, since they have not been encountered except in their 

 descendants which are thus designated. Why this should be so is by no means self- 

 evident, though we may suspect that reproductive isolation of a more effective kind 

 than is achieved by polyploidy may have something to do with it (cf. Manton, 1932), 

 though this cannot be the whole story. The fact, however, forces us to reahze that the 

 fate of a species may depend as much on its method of origin as on any other circum- 

 stance, a conclusion which is perhaps in itself of some importance. 



An observation which emerges far more clearly from the Pteridophyta than from the 

 Cruciferae is the apparently high survival value of the high chromosome numbers. Their 

 accumulation in the most ancient groups was not unexpected, but the tendency of the 

 low-numbered forms to die out first is not so easily explained. That this is the position 

 is, however, suggested in group after group. It is only necessary to recall the numerous 

 cases among ferns {Cystopteris, Polypodium, Dryopteris, Asplenium) in which polyploids are 

 abundant, but the related diploids have to be looked for with care, to be satisfied that a 

 wave of polyploidy is affecting our flora, which may or may not be exceptional and an 

 eff^ect of recent glaciation as discussed on p. 283, but which now gives the impression 

 of a recent replacement of older, low-numbered, species by newer descendants of higher 

 chromosome number. There are exceptions of course, but the really glaring exceptions 

 such as Selaginella (^ = 9), terrestrial Isoetes («= 10), Hymenophyllum (n= 13 and 18) and 

 Osmunda {n = 22), stand out as much by the relative absence of polyploidy as by 



287 



