Abstract: The spore morphology of 20 species of Loxogramme Presl has been studied. The following conclusions may be drawn from the present investigation: 1. In Loxogramme there are two predominating spore types, the trilete and the monolete. 11 species of the material studied belong to the former, and 9 other species to the latter. The trilete spore is circular, subcircular or round triangular in polar view, and ellipsoidal in equatorial view. The polar axis ranges from 26 μ to 65 μ, and the equatorial axis from 33.8 μ to 80.6 μ. The monolete spore is long ellipsoidal or ellipsoidal in polar view, and bean-shaped in equatorial view. The polar axis tangs from 26 μ to 62.4 μ, and the equatorial axis rangs from 40.3 μ to 94.4 μ. The structure of exine of all species studied in this genus is fairly uniform. The outline of spore is slightly undulate or crenate. The exine, about 1.3–3.9μ in thickness and ornamented with small, irregularly shaped verrucae, is usually two-layered, but sometimes its stratification is indistinct. 2. According to the two predominating spore types in Loxogramme, we may divide this genus into two groups: the one belongs to trilete spore, and the other to monolete spore. 3. A few monolete spores were found among the predominating trilete spores of L. assimilis Ching and L. grammitoides (Bak.) C. Chr., while a few trilete spres were found also among the predominating monolete spores of L. parallela Copel, Some spores with varying fissures were observed as well in these three species. According to the degree of variation, two possible “transitional ways” from the trilete type to the monolete type are here suggested. These facts as mentioned above may be significant to the toxonomy of Loxogramma, and may indicate that the trilete type and monolete type are related each other in the evolution of ferns. 4. According to the morphology and anatomy in Loxogramme, Prof. R. C. Ching (1940) separated Loxogramme from Polypodiaceae as a new family， Loxogrammaceae. It has been known that only one spore type i.e. the monolete type is found in Polypodiaceae, while in Loxogramme both the monolete and trilete types are found. It is generally accepted that the trilete is the primitive type, and the monolete is more advanced type. Thus Loxogramme occupies a phylogenetical position more primitive than Polypodiaceae. We agree with Prof. R. C. Ching's viewpoint. 5. The morphology of trilete spore of Loxogramme is very similar to Grammitis Sw. However, Loxogramme is more advanced than the latter from the point of view of palynology, because the monolete type is, as mentioned above, also found in Loxogramme. Thus Loxogramme may be derived from Grammitis as E. B. Copeland (1948) pointed out. It must be noticed, however, that Copeland (1948) put Grammitis in Polypodiaceae, and therefore actually he recognized Loxogramme as a member of Polypodiaceae. R. C. Ching (1940) first separated Grammitis from Polypodiaceae as a new family of Grammitaceae, and only later did Copeland (1951) accept Ching's viewpoint. From the point of view of palynology, it seems that the taxonomic and phylogenetical position of Loxogramme is more closely related to Grammitaceae than to Polypodiaceae.