180 THE BOTANICAL MAGAZINE. rvoi. xxix. No. 346. 



angular lobes, whereas in A. japonica the original three lobes 

 become much deeper and in some cases the central lobe shows 

 a tendency to become again trilobed. The inflorescence of A. 

 japonica is, as Maximowicz describes (14), interrupted, on the 

 other hand that of A. triphylla is continuous. In addition, in 

 A. japonica a short branch sometimes occurs at the base of the 

 spike, as I have already pointed out. Every part of the flower 

 — torus, stamens, pistil — and the fruit is smaller in A. japonica 

 than in the other species. In particular I ma} 7 emphasize the 

 fact that A. triphylla possesses a much larger, broadly flabellate 

 stigma (pi. VII, figs. 17-20, 30-32). One is therefore quite justi- 

 fied in regarding these two plants as entireh' distinct species. 



A. triphylla occurs on the west coast of North America, from 

 British Columbia down to California. A. japonica was first 

 discovered in the Nambu district where it seems to occur but 

 very rarely. It is, however, found in quantities in some parts of 

 Yezo, especially in the neighbourhood of Satporo. According to 

 Komarov (13, p. 324) the same species has also been collected 

 in Northern Corea. 



Some Points of Morphological Interest. 



The Leaf. — The leaf of Achlys, especially of A. japonica, is 

 very variable. In a well-developed leaf of this species the leaflets 

 show a tendency to become deeply trilobed. This is found not 

 only in the terminal leaflet, but also in the lateral ones. In the 

 leaf shown in text-fig. II, 4 the terminal leaflet is deepl}' trilobed ; 

 and the central lobe is again shallowly trilobed and assumes 

 the normal shape of a terminal leaflet. In another case (1) one 

 of the lateral leaflets is very deeply cut into two segments, so 

 that the leaf has become quadrifoliolate. In this example the 

 smaller segment of the divided leaflet resembles a rather feebly 

 developed terminal leaflet (cf. 2). On the other hand a terminal 

 leaflet may happen to be united with one of the lateral leaflets; 

 this results in the formation of a bifoliolate leaf (6). 



These abnormalities suggest that the trifoliolate leaf of Achlys 

 has probably been derived from a biternate leaf. At the same 



