H1LLESHOG, LANDSKRONA, SWEDEN 7 8 PETER BERNSTR()M Two causes of the occurrence of this divergent plant are conceivable. One is that it was the aneuploid whose occurrence was revealed by the cytological examination. For a plant in a hybrid combination which is characterized by a reduced vitality and a distinct physiological unbalance it is not unreasonable to expect that the loss of a chromosome may entail a genetic improvement inasmuch as the chromosome in question has carried a gene, or genes, not harmonizing with the genomes of the opposite parent. The other conceivable cause is that the vitality of the hybrids in this combination of species will depend on what line of L. purpureum has been used. For in the combination L. hybridumX tetraploid L. purpureum the development of the hybrids is in a high degree dependent upon this (ct. below). Thus p7 (like p6) gives dwarf hybrids, while p30 gives vital ones, and this character is Mendelian; more on this point later. The hybrids between L. intermedium and tetraploid L. purpureum during 1944 were obtained after crossing i22 with p7, while those during 1945 -including the aberrant plantwere i47 Xp30. It is probably not necessary to attach any importance to the difference in the mother lines, firstly because L. intermedium has proved to be a very homogeneous species (ct. BERNSTR()M, 1953 b), and secondly because the plant development in hybrids between L. hybridum and tetraploid L. purpureum has proved to depend entirely upon the L. purpureum line used. The plants of p30 that were used for the fertilization of i47 were in part in all probability offspring of spontaneous hybrids with p6 or p7 (ct. BERNSTR()M, 1954). The difference in development as between the hybrid plants of L. intermedium X tetraploid L. purpureum during 1945 may thus very conceivably be due to a different effect of p7 (or p6) and p30. Allohexaploids between tetraploid L. purpureum and diploid 1... amplexicaule may bear out this suspicion (ct. p. 63). Subsequent crossing experiments intended to throw light upon this problem have failed. B. L. intermedium X tetraploid L. amplexieaule a. Cytology It has been possible to analyse cells in I M from four slides of a16 X i22, and reciprocally. As for L. intermedium X tetraploid L. purpureum, it has not been possible to draw any definite distinction between the different kinds of configuration. The results of the studies are summarized in Table 2.Here, as in the case of L. interrnediumXtetraploid L. purpureum, the ANNUAL SPECIES OF LAMIUM 9 10 PETER BERNSTROM another slide with probably 6 III + 3 II + 12 I; it has been drawn to illustrate the difficulty one sometimes finds in distinguishing also bivalents and trivalents. The cell in Fig. 7 is seen obliquely from the side, and in both figures the chromosomes have been separately drawn, with their respective levels retained. The predominant occurrence of trivalents among the configurations and the circumstance that these are very often nine together with nine or somewhat more univalents lead one to the conclus...