S. WALKER 



207 



The three tetraploid species within the complex can have been formed therefore 

 only as a result of hybridisation amongst at least four ancestral diploids, and as the present- 

 day distribution of D. cristata, D, spinulosa and D. dilatata is wide so is the field of search 

 for ancient diploids. 



Hybrids which have been synthesised between the diploids from Europe and Madeira 

 and tetraploid D. dilatata or D. spinulosa are all triploid and show an approximation to 

 equal numbers of bivalents and univalents during meiosis (Table 2). This suggests that 

 the ancestral diploid common to D. dilatata and D. spinulosa is now represented by 

 diploid forms of " D. dilatata " in Europe and Madeira. The diploids from Britain, 

 Norway, Sweden and Switzerland are similar in their morphology and cytological be- 

 haviour; hybrids between any one of them and tetraploid D. dilatata show the same 

 number of bivalents when analysed during meiosis. The diploid from Madeira, however, 

 is distinct in its morphology, being more finely cut than the European diploids and also 

 producing spores smaller in size, which average 42 /u, in length as against an average 

 of 55 fjL. It differs cytologically when involved rn triploid hybrids ; the number of bivalents 

 formed in hybrids with the Madeiran form is constantly lower than in triploids involving 

 the European form; in those with the European form there are 38 to 39 pairs and with 

 the Madeira form 35 to 36 pairs with a possibility of multivalent formation to a very 

 small degree. 



These differences between the diploid forms are significant but nevertheless insufficient 

 to suggest that the diploids represent other than the same ancestral genome from which 

 they evolved separately ; their separation may well have taken place even before the origin 

 of the tetraploid species. That the two diploid forms represent the same ancestral genome 

 can be confirmed by hybridisation between them. This is now in progress. If the two 

 forms are closely related, as seems to be the case from experimental hybrids, chromosome 

 pairing should be high in the hybrid with a possibility of a few univalents or one or two 

 multivalents. 



Separation of the Madeiran form from the normal tetraploid D. dilatata is simple 

 on the type of pinnation alone, although spore size differences apply also as in the separation 

 of the Madeiran and European diploids. It is more difficult to distinguish the European 

 diploid from the tetraploid, particularly in the field. The problem is enhanced by the 

 wide range of form exhibited by the tetraploid itself which may vary in mature size and 

 shape even in the same locality. The most reliable macroscopic characters in the diploid 

 are the marked unevenly deltoid basal pinnae and the greater tendency to tripinnation. 

 However, microscopic characters are more useful and recently Crane (1955) has shown 

 that the spores of the tetraploid and diploid D. dilatata differ in their ornamentation, 

 though not in their size. Other characters are being studied. 



The diploids from Europe and Madeira differ from tetraploid D. dilatata both 

 cytologically and in their morphology. Triploid hybrids, wild and synthesised, are 

 sterile with abortive spores. On this evidence the diploids represent a distinct species 

 and should be separated from D. dilatata. Further, the European and Madeiran forms 

 of the diploid can be separated from one another by their morphology and slight chromo- 

 somal differences exhibited in the synthesised hybrids; they are worthy of subspecific 

 recognition. The final details of nomenclature and taxonomic descriptions of the diploids 

 will be considered in a later paper when the inter-relationships of species in Europe and 

 N. America have been made known. 



The D. spinulosa complex in Europe now consists of four recognisable species, three 

 tetraploid and one diploid, with hybrids between them, both tetraploid and triploid. To 

 what extent, if any, the hybrids are able to back-cross is unknown, though the appearance 

 of mixed populations of D. dilatata and D. spinulosa suggests there may be slight fertility 



