Abstract: The ovuliferous structure of Ginkgo biloba L. has been variously interpreted morphologically. As a result the systematic position and the relationship with other gymnosperms of this ancestral gymnosperm have long been under dispute. In the present paper, a brief survey of the main views as to the nature of the ovuliferous structure is givcn. Based on morphological and teratological data previously reported, a new interpretation is proposed. The essential points are summarized as follows: 1. In morphological essence, a fertile dwarf shoot with some ovuliferous structures in Ginkgo biloba L. might as a whole be nothing but a megasporophyll strobilus (female cone), which is shared actually by all the conifers in the gymnosperms. The fertile dwarf shoot has appearance extremely similar to that of the vegetative dwarf shoot, suggesting that in Ginkgo biloba L. the vegetative organs and the reproductive organs have not been yet well differentiated, and thus its megasporophyll strobilus might represent one of the most primitive compound strobilus types. 2. In Ginkgo biloba, the ovuliferous structure borne in the axil of a scale leaf (sometimes a normal leaf) on the dwarf shoot, together with the scale leaf itself, might be the homogenous organ corresponding to the bract-scale and seed-scale complex of the compound female strobilus of the typical conifers. The complex is a relatively isolated reproductive unit on the strobilus. The normal leaves and the scale leaves on the dwarf shoot might be equivalent to the bract-scales in the typical cones, though the normal leaves still retain the vegetative nature as the foliage leaves on the vegetative shoot. The stalk hearing ovules at its top might be equivalent to a seed-scale of the typical cones. 3. The megasporophyll strobilus in Ginkgo biloba, namely a whole fertile dwarf shoot as mentioned above, seems to show much more primitive characteristics than those of typical conifers. In this plant it is very difficult to distinguish the fertile dwarf shoot from the common vegetative dwarf shoot before reproduction time. Moreover, its megasporophyll strobilus often exhibits more atavistic abnormalities than those of other conifers. All the evidence indicates that the primitive ancestor of conifers might have had the fertile organs which might be of basically identical morphology as vegetative shoots, except that in the fertile organs there might exist numerous fertile leaves bearing one or many ovules. 4. The longer stalk of the ovuliferous structure in Ginkgo biloba might have come from mainly a secondary elongation growth of the seed scale, and only a little part of it might be the remains of the original shoot. The fork structure bearing ovules at the top of the stalk might be the rudementary part of the petioles of the only two extremely reduced megasporophylls. The collar around the base of the ovule might be a secondary protective structure. 5. A correct morphological interpretation of the female strobilus in Ginkgo biloba is doubtless of important significance for our better understanding of the evolution of the female reproductive organs in conifers. According to our interpretation mentioned above, together with the concept of the bractscale and seed-scale complex proposed in the present paper, which is mainly based on the concept of the seed-scale complex propose by Florin, here we put forward an evolutionary theory of the bract-scale and seed-scale complex. According to this theory, the female reproductive organs of the ancestral conifers should be very similar, as mentioned above, to the sterile foliage shoot except that the former might have some fertile leaves which could produce ovules at reproduction time. This ancestral female reproductive organ type might have had evolved towards two directions and thus formed two main evolutionary lines. One is represented by the genus Cycas and we may call it the Cycas Evolutionary Line (C-line), in which the megasporophyll strobilus is monopodial, with the fertile leaves and sterile bracts occurring directly on the main axis. The Cycadaceae is the only living gymnosperm member along this evolutionary line. The second line is represented by all the conifers including Ginkgo, which all have the structure of the bract-scale and seed-scale complex, and thus we called it the "Bract-scale and Seed-scale Complex Evolutionary Line" (BS-line). The members along this line have multipodial female strobilus, i.e. compound strobilus. On the main axis occur some sterile vegetative bracts. In the axils of some or most of the bracts occur the seed-scales. The seed-scales are actually the remains of the extremely, or smetimes completely reduced fertile shoots. Each part of the bractt-scale and seed-scale complex and the main axis of the strobilus could have undergone independent or correlated changes, and thus have had formed various types of strobilus which are found in the living conifers. 6. Our theory on the evolution of the bract-scale and seed-scale complex seems to support the division of all the gymnosperms into two major groups as proposed by Chamberlain, and is also in favour of the placement of Ginkgo biloba into the conifers as the most primitive member along BS-line. 7. Based on their similar morphological characters, it can be considered that Ginkgo biloba might have close relationships with the Nageiaceae, Ephedraceae, Welwitschiaceae and Araucariaceae. All these groups have multinerved leaves without costa. These living gymnosperms might have a more direct relationship withthe ancestral cordaites.

Key words: Ginkgo biloba L, Female reproductive organ, Bract-scale and seed-scale complex, Systematics of gymnosperms