Abstract: In the East Asiatic temperate floristic region, there are many groups of allied plants generally distributed from the Himalayas to Japan or Korea through China. The following spe cies may be taken as examples: Stachyurus himalaicus Hook. f. et Thomas.-S. chinensis Franch.-S. praecox Sieb. et Zucc., Helwingia himalaica Hook. f. et Thomas.-H. chinensis Batal.-H. japonica (Thunb.)Dietrich, Corylop is himalayana Griffith-C. sininsis Hemsl. C. glabrescens Franch. et Sav. It is intriguing to us that the taxa distributed in China are more variable than those in other regions. As considered by Favarges and Contandriopoulus, cytogeography is the only objective method in the understanding of endemics and corresponding (vicariant) taxa. So we believe that the studies on karyological relationship between the Japano Himalayan elements, especially the variation between Chinese taxa in chromosome, morphology and geography (or ecology) will bring some light on the understanding of speciation of vicariants. This paper along with the others (Tang et al. 1983, 1984) is the results of the project “Studies on the Flora of Eastern Asiatic Region”. It is hoped that these studies will eventually deepen our understanding of the origin and differentiation of this flora. The method used here is the same as that in first paper of the series. Voucher specimens are kept in PE. Four species investigated here are cultivated in the Hangzhou Botanical Garden. I. Calycanthus chinensis Cheng et S. Y. Chang (Sinocalycanthus chinensis Cheng et S. Y. Chang) Somatic chromosome number, 2n=22, was determined from leaf-tip cells (pl. 1, fig. 1). The material was pretreated with 0.05% Colchicine solution. The voucher specimen: Y. C. Tang & Q. Y. Xiang no. 84-79. Calycanthus is a genus of only 3 species, C. floridus L. in southeastern United States, C. occidentalis Hook. et Arnott in northern coastal ranges and Sierra Neveda foothills, California, C. chinensis Cheng et S. Y. Chang in Zhejiang, China. The pattern of disjunct distribution of the genus, with two survivors in eastern and western North America and one in eastern Asia shows the genus to be a relict one of so-called Arcto-Tertiary Flora. Doubttless cytological investigation on the 3 species interests us. The chromosome number is n=11 for C. occidentalis (Nicely 1965), n=ll, 2n=22 for C. floridus (Sax 1933, Nicely 1965, for C. floridus var. oblongifolius (Nutt.) B. E. Boufford et S. A. Spongberg (Sax 1933). We come to the conclusion, therefore, that the basic chromosome number of the genus is x=11. 2. Chimonanthus praecox (L.) Link Somatic chromosome number 2n=22 was determined based on mitotic metaphase of leaftip cells (pl. 1, fig. 3.). The material was pretreated with 0.05% Colchicine solution. The voucher specimen: Y. C. Tang & Q. Y. Xiang no. 84-83. The species is widely cultivated and spontaneous only in western Hubei to eastern Sichuan. The present report is in accord with the number reported by Sugiura (1931) and Simonet and Miedzyrzecki (1932). All the records including ours are reported from cultivated plants. Chimonanthus, with 3-4 species, is endemic to China (from the east to the west). The genus is divided into 2 groups. One, with C. praecox only, is deciduous and has sepals and petals rounded at apex. The other, including remaining species, is evergreen or semi-evergreen and has sepals and petals from acute to obtuse at apex. The chromosome number of the species, except C. praecox, is unknown to us. So it is a good material for further study to understand the speciation within the genus. After the brief discussion on Calycanthus and Chimonanthus, it is probably not superfluous to remark cytotaxonomy of Calycanthaceae. Since the establishment of a new genus, Idiospermum based on Calycanthus australiensis by Baker in 1972, the circumscription of Calycanthaceae has been debated. Chant (1978), Takhtajan (1980), Thorne (1983) consider that Calycanthaceae consists of 3 genera (Calycanthus, Chimonanthus and Idiospermum). The subsequent intensive studies on Idiospermum have disclosed numerous differences between the genus and Calycanthus, supporting the establishment of a new family by Walker (1976), Wilson (1976, 1979), Sterner and Young (1980) and Cronquist (1981). No matter what rank is given to Idiospermum, it is indeed related to Calycanthaceae. If Idiospermaceae is merged into Calycanthaceae, it is merely demoted from family rank to a subfamily of Calycanthaceae. So we consider that the discrepancy of its circumscription is not important. The family was sometimes in the past referred to the Rosales. But modern authors, such as Chant (1979), Takhtajan (1980), Cronquist (1981), Thorne (1983), Dallgren (1983), agree that its close relatives are in Laurales. The facts that the family resembles Monimiaceae in a number of features and discontinuous distribution of its members show that the family is of great antiquity. Moreover, the basic chromosome number of the three genera is the same (x=11). It seems reasonable to suggest that the family consists of 3 genera and is subordinate to Laurales. 3. Paris polyphylla Sm. var. chinensis (Franch.) Hara Somatic chromosome number 2n=10 was determined based on mitotic metaphase of ovary wall cells. No B chromosomes were observed (p1. 1, fig. 4). The material was pretreated with 0.05% colchicine solution. The voucher specimen: Y. C. Tang & Q. Y. Xiang no. 84-40. Paris polyphylla Sm., distributed from the East Himalayas to Taiwan Province of China, is a very complex and polymorphic species. Hara (1969) divides it into 3 subspecies (subsp. poly phylla, subsp. marmorata and subsp. fargesii) and nine varieties. We agree with the treatment of Chang (1978), Takhtajan (1983) and Li (1984). They recognize subsp. marmorata and subsp. fargesii as species respectively. Even after these subspecies are separated as species, Paris poly phylla Sm. is still a rather complex one. The range of karyotype variation of the species is as great as that of gross morphology. The species has 8 cytotypes summarized by Hara (1969). Recently Gu (1982) observed 3 different populations of Yunnan province. He has discovered that the numbers of satellites and B chromosomes are variable. Even within a population (from Xundian) two homologous chromosomes of the pair B are different in length. The karyotype of var. chinensis observed by us is somewhat different from those observed by Gotoh & Kikkawa (1937), whose materials are from Taiwan. Besides 2 B chromosomes found in Taiwan plants, the short arms of the homologous chromosomes of the pair C are different in length. Further detail studies to clarify the interrelationship and correlation of cytotypes with morphological characters in the complex are needed. 4. Fritillaria thunbergii Miq. (F. verticillata willd. var. thunbergii (Miq.) Baker) PMC meiotic examination revealed n=12 at diplotene (p1. 1 fig. 2, 5). The material was pretreated with 0.05% Colchicine solution. The voucher specimen: Y. C. Tang & Q. Y. Xiang no. 84-22. The species is distributed in Jiangsu, Zhejiang and Hunan of China and Japan, and also cultivated as a medicinal plant. The present report is in accord with the one by Sato (1942) from Japanese material. A bridge and a fragment were found at AI. Generally considered, bridge and fragment at meiosis indicate the presence of heterozygosity for paracentric inversions. Certainly, heterozygosity for paracentric inversion can result in bridge and fragment, but bridges and fragments may also be resulted from chromosome breakage and reunion. The old literature is reviewed by Newman in the light of these findings and he concludes that the frequency of inversion in plant population has been overestimated (Grant 1975). So the explanation of the configuration observed in this species needs some more valid evidence.

Key words: Calycanthus chinensis, Chimonanthus praecox, Paris polyphylla var. chinensis, Fritillaria thunbergii, Chromosome number, Cytogeography