461 



flowers of the latter, though very small. In Parvatia, as de- 

 scribed and figured by Decaisne, the nectaries (petals) are relatively 

 large in the male flowers, and the nectaries and staminodes are 

 very small in the female flowers. These three examples represent 

 the three proposed genera : Stauntonia, HolboeUia, and Parvatia, 

 the female flowers of which are in other respects exactly of the 

 same structure, differing only in the size and shape of the parts. 

 This reduces the distinctions to the slight modifications indicated 

 in the male flowers, which amount to : 



Presence of nectaries associated with free 



stamens 



« * « 



H 



Presence of nectaries associated with united 

 stamens 



Parvatia. 



Absence of nectaries [from male flowers] ; 



stamens united Stauntonia. 



Mr. Gagnepain does not attempt to show that there are other 

 differences, and, it may be added, there are none in the vegetative 

 characters. Therefore I believe it will be generally conceded 

 that there is no justification for restoring HolboeUia and Parvatia 

 to generic rank. Combined with Stauntonia they together con- 

 stitute a genus as distinct as Deraisnea, Akebia, or Sinofrancheha . 



Coming to species Mr. Gagnepain's views are again opposed to 

 mine, but here it is aggregation instead of segregation, lor 



/< 



XV XK-7 C* w W X v>w*-% va \^r *.■. »■ ■ — » » » » ■ »- - *^ *_* 



HolboeUia latifolia, Wall., he places H. 



forms, as varieties. They are certainly closely allied, but I am 

 still of the opinion that" it is better to call them species and give 



them one name instead of two. 



W 



Hooker's Icones Plantarum.-Part 3 of Volume XXIX of this 

 work appeared in September. It is wholly devoted to the illus- 

 tration of new Chinese and Indo-Chinese species of Impahens, 

 described by Sir Joseph D. Hooker, who has spent the greater 

 part of the' last five or six years studying this difficult genus. 

 The species described and figured exhibit a great variety in floral 

 structure. The author employs the term labelhim for the spurred 

 sepal ; standard for the odd petal, and wings for the two ^-looea 

 lateral petaloid organs, usually considered as consisting ot two 



connate petals. Most 



flowers. I. attopmensis belongs to a group characterised by the 

 lobes of the wings terminating in a thread-like continuation and 

 a strongly keeled, 2-lobed standard. /. Lanessam is one ot a set 

 in which the distal lobes of the two wings are connate and the 

 spur of the lip very short, curved, upwards jmd rtgWj notched. 



verrucifi 



of the blade of the leaf. In I. zygo^nda the sepals are _large and 

 connate, enclosing the very short, broad labellum I. AM> <ts 

 has four weli-devSloped sepals ; a lip with a very broad [spur ■ . d 

 scarcely anv limb ; a very small standard, and very oW £ n °™^ 



I. BrriansaS and /. clanger** ^J^^^Z^ 

 simdar structure. /. LeveUlei is dibtmguisneu uj mi j 



34356 



i 



