106 



P. F. YEO 



which, of the other series of the Ciliatae, they most nearly resemble, but they are relatively 

 well characterized morphologically, and stand apart from the other groups, there being 

 no borderline species. Unlike some groups, the Series Hirtellae does not seem to include 

 species that appear to relate it to more than one other group. In contrast to this, not 

 only do the other Series within the Subsection Ciliatae tend to grade into one another, 

 but the Subsections Ciliatae and Angustifoliae are scarcely discontinuous, members of 

 the Series Alpinae in the Ciliatae approaching quite close to E. salisburgensis of the 

 Angustifoliae. 



The Series Hirtellae, it seems, ought both on morphological and cytological grounds 

 to be raised to the rank of subsection, and the definition of the Subsection Ciliatae amended 

 to exclude them. 



Pugsley (1930) excluded the long-glandular forms known as E. hrevipila var. notata 

 and var. reayensis from the Hirtellae, and this proves to be justified by the cytology. 



Hybridisation 



Although diploid and tetraploid species were reported as long ago as 1932, there 

 appears to have been no consideration of these counts in relation to the hybrids reported 

 to occur. The existence of diploid and tetraploid species imposes a limitation on hybridi- 

 sation and enables members of the two series to exist together, and remain distinct. In 

 fact the results of diploid-tetraploid hybridisation appear to go beyond the formation 

 of an occasional triploid, and it is intended to devote a future paper to this subject. Here 

 it may be added that diploid and tetraploid species commonly grow together and that 

 triploids are rare, that described above being the only one I have found in two seasons 

 of active field work. 



Origin of Tetraploids 



The situation in the triploid E. anglica X micrantha indicates that homology exists 

 between the set of chromosomes present in E. anglica and half the set in E. micrantha. 

 E. micrantha must therefore be an allotetraploid. The simplest inference would be 

 that E. anglica at some time crossed with a distantly related diploid, and that a as result 

 of chromosome doubling in the offspring, E. micrantha arose. However, owing to the 

 close relationship of the various tetraploids, as evidenced by their morphological similarity 

 and the frequent occurrence of fertile hybrids among them, it is possible that an inference 

 of this type may apply to the group of tetraploid species as a whole, and that E. micrantha 

 is a subsequent descendant of the original tetraploid. Similarly E. anglica itself may not 

 have been concerned in giving rise to tetraploids; a related present-day or ancestral 

 species may have been responsible. 



Another possibility is the repeated formation of tetraploid forms from different pairs 

 of diploids from the same two groups. 



It is probable that one should consider the origins of groups of species rather than 

 of individual species. The question arises whether the other diploid group exists today, 

 and if so, where. The only European groups not found in Britain are the Series Pectinatae 

 and Alpinae; these fall within the Ciliatae and include species which form a series linking 

 the Ciliatae with E. salisburgensis in the Angustifoliae. It is therefore unlikely that they 

 include diploid species. Within the Angustifoliae, however, there is a sharp discontinuity 

 between the salisburgensis complex, and the two closely related species E. cuspidata Host 

 and E. tricuspidata L. A chromosome survey is needed, including these two species, 

 and covering the other subsections of the Section Semicalcaratae, namely the Alpicolae 

 and ]aponicae Pugsley (1936) of Japan, and also the Section Atlanticae Pugsley (I.e.) of 

 the Azores. It is doubtful if the Alpicolae and Japonicae are very distinct from one 



