174 



THE BOTANICAL MAGAZINE. 



[Vol. XXIX. No. 346. 



and Ito (11, p. 436) on the other hand state that the spike of 

 A. triphylla is interrupted. This statement is certainly incorrect. 

 When young, the spike appears to be looser near the base and 

 much more compact towards the apex. This is due to the fact 

 that the apical region is much younger than the basal part. 

 Hooker's figure (10, tab. 12) shows a few of the basal, more 

 mature flowers widely separated from the rest, but this is really 

 clue to exaggeration on the part of the delineator and is very 

 misleading. In A. 

 japonica the spikes 

 are, as a rule, in- 

 terrupted particular- 

 ly in their basal 

 region, a few small 

 groups of flowers 

 being distantly dis- 

 posed on the rhachis. 

 The actual arrange- 

 ment of the flowers 

 on the rhachis can 

 be accurately ascert- 

 ained in a deflorate 

 or fruit-bearing 



spike, in which full 

 growth of the 

 rhachis has already 

 been attained 1 ' (cf. 

 text-fig. Ill, 2 and 

 3). 



A very interest- Text . fig m 



ing fact to be noted Achlys japonua Maxim, xf. 



is that in A. japonica In 2 and 4 lhe P lei °chasial branch is very 



n < ' n , _ clear, while in the others it is represented by 



a small branch (5- . f . „ ,, . „ . . . J 



an imperfect flower (in 1 a flower just above 



lo mm. in length) this imperfect flower is more conspicuous). 



1) The rhachis of ^4. triphylla measure.-, in fruit-bearing stage, 3-8 cm. in length. 



