ON SOME POINTS IN THE MOKPHOLOGV OF HALOPHILA. 291 



the genus — we observe a procumbent, freely branched, jomted 

 stem : from almost every joint arises a branch, jointed and pro- 

 cumbent, like the axis fi'om which it springs, and it in tarn bears 

 similar branches, w^iich repeat the i3rocess. At the joints there 

 is visible a pair of small sheathing scale-leaves, one on the 

 under, and the other on the ux3per side of the procumbent stem ; 

 and in addition there is a i^air of foliage-leaves on the uj)per side 

 of the stem, with their bases somewdiat embraced by the sheathing 

 upper pair of scale-leaves. The arrangement is such that at first 

 sight one might suppose we had stipulate leaves ^jroduced in pairs 

 at successive joints on one x3rocumbent branching axis. 



But the arrangement is more complicated and, so far as I 

 know, unique in the vegetable kingdom ; and it is to this — a x^oint 

 briefly noticed by Irmisch" many years ago — I wish, in the first 

 l^lace, to direct attention. 



Taking any one procumbent axis as the (relative) j)rimary axis, 

 we find it consists of alternating elongated and shortened inter- 

 nodes. Each internode bears a small sheathing scale-leaf, and 

 as every alternate internode is shortened these scale-leaves are 

 approximated so as to appear in pairs. In each pair the leaves 

 are always jDlaced opposite each other ; the older, that of the 

 inferior internode, bemg always on the under side of the stem. The 

 successive pairs are not superposed, as has been hitherto believed, 

 but are set on so that their median planes form an acute angle 

 with one another and a less acute angle with the vertically bisecting 

 plane of the primary axis. The (relative) primary axis, no matter 

 how greatly it elongates, only bears those scale-leaves. 



In the axil of the uppermost (younger) of each pair of scale-leaves 

 on the (relative) primary axis there arises an extra-axillary leaf-bud, 

 always on the side of the median plane of the subtending leaf, 

 farther from the vertically bisecting plane of the primary axis. 

 The secondary axis so developing begins with two shortened 

 internodes, each of which bears a foliage-leaf. These foliage- 

 leaves are always opposite one another, and are necessarily closely 

 approximated on the axis. The third internode of the secondary 

 axis is elongated, and bears a scale-leaf; and the fourth internode 

 is shortened, and also bears a scale-leaf, these being placed 

 opposite one another. Thus at the joint formed by' the third 

 and fourth nodes a pak of opposite scale-leaves occurs. Now 

 it is curious to note that these scale-leaves are in series with the 

 scale-leaves on the primary axis ; and at every succeeding joint on 

 the secondary axis a like pair of scale-leaves may be produced, and 

 successive pairs will be set on at an acute angle with the pre- 

 cedmg. We have, in fact, a repetition of what we observed on the 

 primary axis. The upper scale-leaf of each pair on the secondary 

 axis gives rise to an extra-axillary tertiary axis, the first pair of 

 leaves on which are foliage-leaves, and all succeeding pahs are 

 scale-leaves, and so on; 



Now whilst the scale-leaves on aU the axes are in series, the 



* Bot. Zeit., xvii. (1859), .'{55 adnot. 



