Occurrence of Folia gc-kavcs in Ruscus anclrogynus. 133 



marked, while tlie petiole is shortened and the lamina 

 much reduced. 



When these foliage-leaves are compared with the cladodes 

 (developed on the aerial stem), the following striking 

 differences may be noted. 



1. In vernation the foliage-leaf is convolute, while the 

 cladode — if I may speak of " vernation" in this connection 

 — is perfectly flat. 



2. The foliage-leaf is provided with a long stalk, while 

 the cladode is nearly sessile. 



3. The foliage-leaf exhibits a distinct midrib, while the 

 cladode does not. They resemble each other, however, in 

 the longitudinal disposition of the veins, of which every 

 fourth one is stronger than the intermediate ones. 



4. In the foliage-leaf, the non-stomatic surface is the 

 upper, and the stomatic the lower, — as in ordinary leaves. 

 In the cladode, on the other hand, the non-stomatic surface 

 is morphologically the lower, while the stomatic surface is 

 morphologically the upper (Plate X.) ; these surfaces, how- 

 ever, becoming inverted by a twist at the base of the 

 cladode, whereby the stomatic surface is directed down- 

 wards and the non-stomatic upwards. In this arrange- 

 ment, R. anclrogynus resembles B. [Dcmde) racemosus, and 

 differs from B. aculeatus, where the stoniata are equally 

 disposed on both surfaces of the cladode, which is twisted 

 only so far as to direct its edges upwards and down- 

 wards.'*^ In the allied Myrsiphylhim asparcigoicles, the 

 stomata are developed exclusively on the morphological 

 lower surface of the cladode, which is not twisted at all. 



The almost exclusive development of the stomata on the 

 morphological upper surface of the cladode, and the inver- 

 sion of the surfaces by a twist, in R. anclrogynus and R. 

 racemosus, is, in many respects, of great interest. A precisely 

 similar phenomenon is to be noted in the leaves of the 



* Goebel (Schenk's Ilandb. d. Botanik, iii. 1, p. 269) refers to R. raccvioms 

 and R. aeuleatus as both having their cladodes twisted about 90°, so as to 

 direct their edges, instead of their surfaces, upwards and downwards, like 

 the phyllodia of New Holland Acacias ; a statement correct as regards R. 

 aculcafus, but not so as regards R. racemosus. In R. androgijnus, the amount 

 of twisting varies with the position of the branch supporting the cladode, 

 being just sufficient to effect the downward direction of the stomatic surface, 

 — or, rather, its direction away from the light. In R. Hi/po/jlussujii, although 

 the stomata are equally distributed over both surfaces, there is practically no 

 twisting of the cladoih'. 



