Discussion

Cissus quadrangularis

The morphological characters especially of stem, leaf, tendril and season of flowering of C. quadrangularis are described differently in different Floras and treatises of medicinal plants. The existence of variants in this species is not recorded in many botanical literatures. Our observation in different seasons and localities on the presence of three variants prove beyond doubt that the Variant I has four sided and winged or margined stem. Variant II has four sided but wingless stem and Variant III has flat stemmed (two sided stem). Earlier workers observed that stem of the species is nearly leafless or leafless when old (Hooker, 1978 and Kirtikar and Basu, 1980). We noted after careful observation in different seasons that in all the three variants new leaves and branches appear only during rainy season. Leaves being caducous, the stem appears almost leafless in other seasons. The nature of tendril in C. quadrangularis is variously described as forked or not, slender or stout. (Matthew, 1983 and Kirtikar and Basu, 1980) In our observations, forked tendrils are present occasionally in Variant III (Fig. 5B) and unbranched tendrils in other two variants. The tendril is long and slender when young but strong and hardy when mature. The presence of a node with a bract in the middle of the tendril in all the three variants has been noted in the present work. Difference in the observations on the periodicity of flowering in C. quadrangularis may be due to the observations of specimens in different geographical locations. All the three variants of C. quadrangularis grown wild and in herbal gardens in this locality (Thanjavur) are flowering during the period of April-July. The existence of three variants among C. quadrangularis was not recorded in earlier Floras. Kumbhojkar et al. (1991) noted the presence of two varieties, but he regarded them as 'cultivars'. Our continuous observation and after prolong cultivation by stem cuttings of the three variants proved that the three variants bred true. However, in Variant Ill, occasional formation of three sided internodes was observed. Warrier et al. (1994) also observed the formation of three sided branches from the four-sided variety. As observed by kumbhojkar et al. (1991), the wingless and two sided varieties (Variant II and Variant Ill) could not be assigned as 'cultivars', since Srinivasan of Botanical Survey of India, Coimbatore, collected the wingless variety of the species from wild and the same is deposited in Madras Herbarium. BSI, Southern Circle, Coimbatore.

Anatomical characters of stem, petiole, leaf and tendril of the three variants of C. quadrangularis do not show marked variations. Presence of pearl glands reported in earlier works (Metcalfe and Chalk, 1957 and Janardhanan et al. 1981) was not observed in the specimens, of the present work. However, the presence of ephemeric exudates is reported on the surface of the aerial parts of the three variants of C. quadrangularis for the first time. Janardhanan et al. (1981) observed abnormal Stomata and cuticular striations on leaf epidermal cells. No abnormal stomata are observed in the present work in all the three variants. However, cuticular striations are noted occasionally in Variant I only. The anatomy of the stem of Variant I is similar as observed by Madan and Nair (1959). No secretary cells are observed in the stem but presence of mucilage cells is noted in the present work. The variants of C. quadrangularis reveal more similarity than dissimilarity in physicochemical and phytochemical studies. Vitamin C content is maximum in Variant III among the three variants studied. Based on the present observation, it is suggested that Variant II and Variant III might be the chimeric variations of variant I. Atal et al. (1978) found that plants of C. quadrangularis excluding roots are devoid of tannins. In the present observation, alcoholic and water extracts of tender shoots of Variant I gave positive result for tannins when lead acetate is used as identifying reagent. Alcoholic extract of Variant II & III also gave positive result with ferric chloride and gelatin solution. Lead acetate solution gave negative result to Variant II but positive to Variant III. Water extracts of Variant II & III show positive results with gelatin solution and lead acetate solution but negative in ferric chloride solution.

Madhuca longifolia

Our present observations on the morphology, anatomy and physicochemical standards of M. longifolia var. latifolia and M. longifolia var. longifolia slightly deviate from the previous reports. We observed that the color of the bark of var. latifolia is dark brown to black and that of var. longifolia is brown. However, according to earlier reports (Anonymous, 1962 and Kirtikar and Basu, 1980) color of bark of val. latifolia (Syn. M. latifolia and M indica) is grey (Gamble, 1967) or grey to dark brown (Anonymous, 1967) or dark brown (Kirtikar and Basu, 1980). Our present observation on the numbers of anthers present in var. latifolia is 16-18 and in val. longifolia is 18-20. Conforms to the earlier observation of Mukerji, 1953, and Kirtikar and Basu, 1980, However, Matthew, 1983 reported the number of anthers in var. latifolia as 16 and in var. longifolialongifolia is strictly one seeded and var. latifolia is 1-6 seeded. It is to be noted that the observation of 6 seeds/fruit is the first report. The above minor differences in the observation on the morphology of bark, number of anthers and number of seeds/fruits might be due to the variations in the observed samples and due to the climatic and geographical variations exist among different populations of the species. as 18. Difference in the observation of number of seeds/fruit also" exists between the present observation and previous reports (Mukerji, 1953, Matthew, 1.9813 and Kirkitar and Basu, 1980). According to present report var.

Anatomical characters of plant parts of two varieties of M. longifolia observed in the present work are identical with the previous reports made by Metcalfe and Chalk, 1957. We observed for the first time the presence of extended prongs and rods in the cuticle of abaxial epidermis in both varieties. This character may be considered as advancement towards the attainment of the pappillose cuticle of Diospyros sp. of Ebenaceae. Atal et al. (1978) reported the stem bark of M indica is devoid of tannins. However, in the present observation tannins are quantitatively estimated in both the varieties of M longifolia.

In the previous reports, oil yield from seed kernels is estimated to be 40-45% (Anonymous, 1962) and 50.55% (Mukerji, 1953). However, in the present work highest oil yield (59.77 %) is reported in var. latifolia. The population of var. latifolia from which the study was made would form a high yielding elite group. Further selection procedures are needed to select high oil yielding strain of val. latifolia. The content of invert sugar in air-dried corolla of var. longifolia and in var. latifolia was observed as 55.88% and 47.75% respectively in the present work. Mukerji (1953) observed 52.6% invert sugar in the flower of 'Mahua'. However, study samples in the earlier study consisted of a mixture of both varieties of M. longifolia and M. butyracea. The anatomical, physicochemical, phytochemical and analytical standards of oil are showing more similarity between the varieties but quantitatively they show marked differences. In general, these two varieties are closely related with one another as stated in the previous reports (Anonymous, 1962). But medicinally val. longifolia is used in South India and var. latifolia is used in North India (Warrier et aI., 1994).

Basella alba

Previous authors considered both green and pink color specimens of B. alba as single species (Gamble, 1967, Hooker, 1978 and Matthew, 1983). Recently, Henry, et ai. (1987) raised the two specimens to variety level as B. alba var. alba and B. alba var. rubra. Our present breeding experiments on the two varieties proved that the color of var. rubra (presence of anthocyanins pigments) is genetically fixed. The present pharmacognostical work on the two varieties of B. albaB. albarubra. Anatomical studies conform to the earlier reports (Metcalfe and Chalk, 1957 and Sharma, 1961). Metcalfe and Chalk (1957) reported only the presence of paracytic stomata. However, in the present work both paracytic and anisocytic type of stomata is found. Also, there is no significant difference in the analytical values and phytochemical tests of the two varieties. Atal et al. (1978) reported that the tannins are absent in 50% alcoholic extracts of aerial parts of B. rubra. However, in the present observation it is found that the tannins are present in alcohol and water extracts of both the varieties of B. alba. The two varieties of B. alba, viz, var. alba and var. rubra of the vegetative and floral parts. revealed that there is no anatomical difference among them, except purple pigmentation of var. show more similarities than dissimilarities in the anatomical and physicochemical characters. The only significant variation is the color