Section 8 — Cytatoxonomy and Experimental Taxonomy 



plastids occurs in the zygotes, and the subse- 

 quent plastid competition will lead to a gradual 

 elimination of the plastid type which multiplies 

 more slowly. It is a proven fact that this process 

 did happen during phylogeny, since wild species 

 which have been shown to be compatible with 

 more than one plastid type, now possess the 

 faster multiplying plastid type. 



3. Sometimes the incompatability between 

 certain genomes and plastomes is expressed even 

 in the microspores by inhibited germination of 

 living pollen grains or by early cessation of 

 pollentube growth. As this character shows up 

 mainly after hybridization, Renner (1919) 

 regarded it as the mechanism responsible for the 

 elimination of one complex in the pollen. This 

 mechanism would tend to make the complex- 

 heterozygotic species heterogamous. Species of 

 the parviflora-groups would seem to offer 

 examples for this mechanism. Furthermore, 

 based on experimental results about plastid com- 

 petition and about the compatibility correlations 

 between the five wild type plastids and the 

 genotypes of the subgenus Euoenothera, the 

 phylogeny of the five plastomes of this subgenus 

 are being inferred step by step in an unambiguous 

 manner. 



8.3. Chromosome Differentiation in Genomes of 

 Gossypium. M. S. Brown (College Station, 

 U.S.A.). 



Chromosome studies in several leguminous 

 genera made by various investigators suggest 

 heterogeneity within the genera as to base num- 

 ber, chromosome dimensions and shape. 



Investigations in Indigofera from West and 

 East Africa, and in a few Himalayan and East 

 Asiatic species, reveal differences in base number, 

 shape and size between the various species, and 

 a few cases of polyploidy. 



Recent data for the genus Vigna show two 

 base numbers, 10 and 1 1, this in accordance with 

 earlier reports, and considerable differences in 

 chromosome dimensions within the Vigna 

 unguiculata complex. Grafting experiments be- 

 tween some cultivars of the cultivated cowpeas 

 resulted in a success of 80-90 per cent when 

 using types with equal chromosome dimensions, 

 and a success of below 50 per cent in the case of 

 deviating dimensions. 



There exist several other genera, e.g. Trifolium, 

 Astragalus, Lespedeza, in which either ranges of 

 base numbers or entirely different base numbers 

 are present. Sometimes such a difference agrees 

 with different gene centres. In some case loss or 

 increase of chromosomes has been suggested. 



Another line of thought next to those men- 

 tioned above may be the acceptance of poly- 

 phyletic evolution: the same lines of develop- 

 ment acting on different original substrates under 

 similar circumstances, but at various points of 

 time may result in convergent morphological 

 types brought under an equal denomination by 

 the taxonomists. 



Chromosomes of species and Fi hybrids of 

 Gossypium are compared at pachytene, diplotene, 

 diakinesis and metaphase. At pachytene, all 

 chromosomes pair closely, even in sterile hybrids 

 between species with chromosomes of unequal 

 size and low bivalent number at metaphase. At 

 diplotene, genome differences become apparent; 

 and at diakinesis and metaphase, size differences 

 in chromosomes are shown to be correlated with 

 amount and distribution of chromatic material. 

 In monosomies of Gossypium hirsutum, an allo- 

 tetraploid, chromosomes of A and D genomes 

 can be distinguished in like manner. These 

 differences in chromosome size and composition 

 can be correlated with variation in the cytosine/5- 

 methylcytosine ratios as determined in A, B and 

 D genomes by Ergle and Katterman. 



8.4. Signification of Cytological Data for the Tax- 

 onomic Conception of the Leguminosae. J. A. 



Frahm-Leliveld (Wageningen, The Nether- 

 lands). 



8.5. Desynapsis in Lotus Hybrids. William F. Grant 

 (Quebec, Canada). 



In meiotic analyses of several embryo cultured 

 interspecific hybrids between six diploid (2n = 

 12) species of Lotus (Leguminosae), closely 

 related to L. corniculatus, univalents, resulting 

 from precocious separation of the bivalents, were 

 observed at diakinesis and metaphase I. In some 

 cells up to 50 per cent of the chromosomes were 

 observed as univalents. Univalents, however, 

 were present in only 40 per cent of the cells 

 examined. Desynapsis is considered to be pri- 

 marily of the "weak" type as defined by Prakken. 

 In some cases, at metaphase I, the univalents were 

 observed to be loosely connected. At anaphase 

 I, lagging chromosomes were observed, most of 

 which consisted of univalents. Occasionally 

 lagging chromosomes were observed to remain 

 behind in the cytoplasm at telophase I and to fail 

 to be incorporated within the restitution nuclei. 

 In some cases the telophase I chromosome com- 

 plements consisted of more, or less, than the 



132 



