APOGAMOUS FERNS. EVOLUTION OF THE SEPARATE SPECIES 



is not a negligible possibility is made clear by the behaviour of the two upper members 

 of the series (the tetraploids and pentaploids), which also provide such genetical evidence 

 as we possess. That they are hybrids between the lower numbered forms of D. Borreri 

 and the Male Fern is suggested not merely by their isolated occurrence, their chromo- 

 some number and their mixed morphology, but also by a curious circumstance affecting 

 the total spore output. This is always low, as may perhaps be seen in Fig. 201, in which 

 the relatively good spores of diploids and triploids contrast quite strongly with the high 

 proportion of shrunken spores found in tetraploids and pentaploids. So marked is this 

 character that inspection of the spore output of an unknown plant in which hybridity 

 of this kind is suspected can provide almost as reliable a guide to its chromosome 

 number as in other cases of hybrids in which apogamy is not involved. In these plants 

 there are, however, always a few very large good spores produced which, on sowing, 

 give a sparse crop of apogamous prothalli which reproduce the parental type exactly. 

 The low spore output is therefore presumably the reason why homogeneous local 

 populations are not formed. 



The reason for the low spore output is at once to be seen on sectioning a sorus. Where- 

 as in diploid and triploid (with the exception of var. polydactyla Dadds) the eight-celled 

 type of sporangium is so conspicuous as to dominate the general field, in tetraploids and 

 pentaploids it is the sixteen-celled type which predominates to the extent that in one 

 tetraploid for which accurate counts of the two types were made, sixteen-celled out- 

 numbered eight-celled by 17:1. It is, therefore, a rather exceptional sporangium 

 which is able to carry through the normal apogamous development, and viable spores 

 are therefore correspondingly scarce. 



This is a very curious type of genetical expression of a character. Apogamy, though 

 inherited, is not behaving at all like a simple Mendelian dominant, nor is it even a 

 partial dominant in the usual sense. Combination of an apogamous with a sexual 

 species might perhaps be expected to produce the intermediate type of sporangium 

 described as type 3 on p. 166 if dominance were imperfect, yet we find merely a 

 reduced percentage of eight-celled sporangia in which, however, the apogamous 

 qualities are fully developed. 



This type of inheritance, it may be said in passing, provides strong confirmation of 

 the hybrid origin of the tetraploids and pentaploids under discussion (and perhaps also 

 of some triploids such as var. polydactyla Dadds), since it agrees exactly with what has 

 been found in a hybrid of known origin described and synthesized by Dopp. Dopp's 

 hybrid was between a diploid crested form of D. Borreri and normal D. Filix-mas, and it 

 must therefore have been tetraploid. The crested character proved to be recessive, 

 though the hybrid resembled D. Borreri in the shape of the pinnae and indusium. It 

 also resembled D. Borreri by reproducing apogamously, though the spore output was 

 low owing to a preponderance of sixteen-celled sporangia. 



These facts are not only interesting in themselves but they provide an opportunity 

 for testing the possibility that occasional viable spores may be formed in the sixteen- 

 celled sporangia. An experiment of this kind was carried out on a pentaploid from 

 Ireland. This was lifted one autumn and transferred in a large pot from the mixed 

 population of the experimental garden to an isolated greenhouse about a mile away on 



MFC ig2 ^^ 



