Takeda — Old World Species of the Genus Mahonia. 



terminal extension of rhachis cannot be taken a 

 for distinguishing M. Fortunei and M. confusa, the above- 

 mentioned character is always decisive.* 



It hardly needs mentioning that the leaflets are sessile among 

 the species belonging to the " Longibracteatae."t 



The inflorescence is a raceme, usually simple, but occasionally 

 branched near the base, thereby more or less assuming the nature 

 of a panicle. The flowers are invariably stalked and are disposed 

 on the rhachis either rather loosely or densely, according to the 

 species. They are arranged in a spiral, or in some cases even in 

 a subwhorl.J 



The bracts subtending each flower afford a good distinguish- 

 ing character of species. They may be small, scaly, pointed or 

 blunt, narrow or broad, and so forth. They are persistent, and 

 increase in size at the time of fruit, but as a rule only very 

 slightly. 



The pedicels may be longer or shorter than the bracts 

 which subtend them. They may in the fruit-bearing stage 

 become thickened, recurved, spreading, or nodding only at 

 the apex. 



The prophylls are as a rule entirely suppressed in the group 

 Longibracteatae.% It is however of much interest to record here 

 that conspicuous prophylls, similar to, or sometimes smaller and 

 narrower than, the bracts are seen in our new species, M. 

 bracteolata and M. Mairei. They are present in most of the 

 flowers of these species situated below the middle of the in- 

 florescence, occurring one or two on each pedicel. 



The flowers are yellow in colour, and are, in some species, 

 fragrant. Three whorls of sepals are invariably present, two 

 whorls of petals, two whorls of stamens with three members in 

 each, and a pistil in the centre. || A few species have however 

 been described as having only six sepals. Examination of 

 specimens has convinced the present writer that those statements 

 are erroneous. In most cases the sepals of the outermost whorl 

 are minute and much smaller than the others. In some cases 

 however they are almost or quite as large as those of the 

 middle whorl or even of the innermost whorl. To regard the 

 outermost sepals as bracts is absurd.^ 



Shape, size, and nervation of sepals and of petals are liable 

 to variation to a certain extent, vet they exhibit very distinctive 



t Cf. Fedde. 



X Fedde, I.e. 



unequal growth ■ 



