1. The initial work on hybrids betweenLolium perenne andFestuca pratensis, reported in the original paper of the present series (Jenkin, 1933), has been followed up. 2. Further initial crosses gave results in good general agreement with those already reported with one important exception. 3. The exceptional.Fb1 hybrid plant was produced when a new breeding procedure was adopted on an experimental basis. This plant was of full normal vigour even at germination. 4. It is argued that this exceptional plant was the result of a special type of fertilization which may also occur in nature. 5. The reported prevalence ofF. loliacea in some areas can then be satisfactorily explained. 6. The anthers of the artificially producedFb1 hybrids are non-dehiscent, and the female fertility of the plants is very low. 7. Some of theFb1 hybrid plants produced artificially would be immediately classifiable asF. loliacea, but others would present some difficulties in classification. 8. Peto (1933) had previously shown that naturally produced plants classifiable asF. loliacea might be either diploids or triploids. 9. It would not be possible with certainty to classify either the artificially producedFb1 hybrids or the naturally producedF. loliacea plants into diploids and triploids either on morphological or on growth form characteristics because of the possibility that the differences already existing between the parent plants might overrule the differences due to diploidy and triploidy. 10. The general similarity between naturally occurring diploidF. loliacea plants and some of the diploid artificially producedFb1 hybrids supports the contention of those taxonomists who have held the view, based on other considerations, that diploidF. loliacea plants at least are the result of natural intercrossing betweenL. perenne andF. pratensis, but the natural triploids might still be different in origin. 11. Little progress could be made by using the pollen ofF. pratensis on theFb1 hybrids, but one established backcross plant of this type was obtained. The present report is therefore mainly based upon results obtained by usingL. perenne except where fertile hybrid derivatives were used. 12. From the study of second-generation plants, it is concluded that the artificially producedFb1 hybrid plants at gametogenesis gave rise to at least three types of effective female gametes. 13. A proportion of the backcross (toL. perenne) hybrids were effectivelyL. perenne plants of high fertility, although they may have carried someF. pratensis genes. 14. Other backcross hybrids of the same derivation were of relatively very low fertility, although morphologically they also approachedL. perenne rather closely. It is suggested that low fertility in these plants was due to interchange of chromosome sections at meiosis in theFb1 hybrids. 15. One of the backcross hybrids was a triploid (Peto, 1933). The possible origin of its triploidy is discussed. 16. Further breeding work with diploids to produce third and later generation plants suggests that genes derived fromF. pratensis may be transferred to and carried by derivatives which are effectivelyL. perenne plants. 17. The artificially produced triploid backcross hybrid plant that was extensively studied was to a slight extent (and effectively) both, male and female-fertile. 18. The derivatives of this triploid plant showed a wide range of variation in type even within a single cross withL. perenne. It is suspected that one plant of this family and another from a cross with the hybrid derivative plant 23-bE-7/l were approximateL. perenne autotetraploids, but no cytological evidence is available. 19. A near-triploid plant, 23-bE-7/l, is brought into the discussion on txiploids, although it was not derived fromFestuca pratensis. 20. Using data derived from Peto (1933), Thomas (1937) and Myers (1944) an attempt is made to trace the cytological history of the known triploids and near-triploids studied. 21. It is then suggested that at gametogenesis in theFb1 hybrid parents of the two plants 56-bE-2 and 23-bE-7/1 crossing-over with subsequent interchange of chromosome sections occurred. 22. On the average figures given by the authors quoted there is no evidence that this had happened in the history of the natural triploids. It is therefore suggested that these plants were first-generation hybrids, produced either (a) by intercrossing betweenL. perenne and a tetraploidFesluca type which is not yet identified in Britain or (b) betweenL. perenne andF. pratensis by a special type of double fertilization. 23. The possible effect of intercrossing betweenL. perenne andF. pratensis and the subsequent production of backcross and other derivatives of such hybrids upon a natural plant population is discussed.