Detailed study of the structure and behaviour of chromosomes in the somatic tissue of twenty-one species of the genusDracaena has been made and the “normal” chromosome numbers of all these species have been reported for the first time in this paper.

The previous and the present records suggest that most of the species possess a number of chromosomal biotypes. These biotypes mainly differ with respect to their chromosome numbers.

Relationship between species showing multiples of different series of chromosome numbers, viz. eight, ten, thirteen, seventeen and nineteen has been indicated.

On the basis of the fact that a general resemblance in gross morphology of chromosomes and similarity in total amount of chromatin length are present amongst different species of the genusDracaena, it has been suggested that the species of this genus represent a homogeneous assemblage in spite of the fact that inconstancy in chromosome number is noted within a species.

The different lines have mainly been assumed to have come out through continued production of aneuploid numbers during evolution. The presence of a number of chromosomal biotypes indicates that such aneuploid numbers often arise.

Minor differences in details of chromosome morphology, and the presence of super-numerary constrictions in certain species, have been regarded as proving that structural changes of chromosomes have also played a distinct role in evolution of the species.

As the different species ofDracaena are propagated exclusively through vegetative means, the only explanation for the origin of biotypes which can be suggested is that the recorded variant nuclei enter into the formation of new daughter shoots from which new individuals originate with different genomic constitutions. As flowers are scarcely noted and sexual reproduction is entirely ineffective with respect to propagation, this seems to be the only way through which speciation is effected here.

The present investigations on the chromosome study of ten Indian representatives ofViola have revealed interesting data from cytotaxonomical and cytoecological standpoints. The existence of two distinctly different chromosome series in the same species, rather uncommon for this genus, has been recorded inV.serpens (2n= 18 and 48) andV. biflora (2n = 12 and 20). Such intraspecific variations imply that differences in numerical series of chromosomes should not necessarily play an important role in categorizing species assemblages. They have also been noted inV. tricolor (2n = 42 and 46) and polyploid cytotypes have been recorded inV. hookeri (n = 9 and 36). On the basis of chromosome data, it has been stated that the retention ofV. odorata andV. tricolor under the section Erpetion as followed by Engler and Prantl, is not desirable. De Candolle in his system kept the two genera under two separate sections. Remarkable similarity betweenV. serpens ( 2n = 18 and 48) andV. distans ( 2n = 44) in chromosome morphology has been shown.

From a study of the chromosomes of different ecotypes it has been shown that inViola extreme climatic conditions are tolerated by diploids, whereas the polyploids abound in comparatively moderate climate, which is rather uncommon for other genera.