210 Takeda— Old World Species of the Genus Mahonia. 



M confusa* This species— M. confusa— is moreover described 

 by Schneider in his work as M. Zemanii, as the present writer 

 has pointed out.t As a matter of fact, there are a few others to 

 which no references at all are made in Schneider's enumeration, 

 some long known from India, and some from China.J 



Before we enter into the systematic account of the genus 

 some remarks may perhaps be made upon morphological 

 features.§ 



The leaf of Mahonia is always impanpinnate and is furnished 

 with a pair of stipules at the base. The stipules are adnate to 

 the petiole and only their upper free portions are noticeable. 

 For the sake of simplicity, these free portions alone are taken 

 into account in the descriptions given in the following pages. 

 They may be long, short, deflexed, curved, straight, and so forth. 

 When the stipules are completely united with the petiole, they 

 are conveniently described as being " absent." 



The rhachis of the leaf extends, as a rule, beyond the upper- 

 most pair of leaflets, so that the terminal leaflet appears as if 

 " petiolulate." The length of this terminal extension of rhachis 

 varies to some extent, even within one species. In certain 

 species the terminal extension is usually reduced to a minimum 

 in length, resulting in the terminal leaflet becoming apparently 

 " sessile," e.g. in M. Fortunei. Sometimes a terminal leaflet is 

 united at the base with one of the leaflets of the uppermost pair 

 which often appears to consist of a single leaflet. In this case 

 the terminal leaflet is broader than usual and is more or less 

 deeply cleft on one side. In rarer instances such conspicuous 

 lateral incisions, which indicate fusion of two or more organs, 

 cannot be detected in the terminal leaflet, consequently the 

 leaflet is only abnormally broader and apparently more lengthily 

 " petiolulate " than usual. The presence or absence of the 

 terminal extension of rhachis cannot therefore be regarded as a 

 good distinguishing character of species in this genus. In the 

 following descriptions no mention is made of the presence of this 

 extension of rhachis unless it happens to be of diagnostic value. 

 In this connexion it may perhaps be pointed out that the length 

 of petiole, namely, the distance between the base of the petiole 

 and the lowermost pair of leaflets, is an important diagnostic 

 feature not to be overlooked. || In many species the petiole is 

 very short, measuring only a few centimeters, whilst in some it 

 is comparatively long. While the presence or absence of the 



* Sprague, in Kew Bull., 1914. P- 2 32. 



t Takeda, ibid., 1915, p. 128. 



J For example, M. acanthifolia, M. Leschenaultii, M. trifurca, M. ganpiensis, 



§ In referent 



