532 ME. G. C. BOURNE ON THE (lENUS LEMNALIA AND 



and a, h, c, d, ^.y (lateral members) were traced into such a branchlet, of whicli a section 

 is represented in fig. 20 (drawn to a larger scale than fig. 19). It is seen that, whereas 

 the three chief members retain their predominant size and occupy the bulk of the section, 

 the lateral members of the branch have diminished in number. Thus a is emerging as a 

 small authocodia ; h and e have disapjieai-ed, having come to the surface in similar 

 fashion lower down on the branchlet ; c, d, and /"have increased in size relatively to the 

 chief members and will emerge as anthocodite higher up on the branchlet. The two 

 lateral membei-s marked .c and i/ were not jH'esent in the branch. They took their origin 

 from solenia growing out of the chief members in the branchlet after its separation from 

 the branch, ,v arising from A and >/ from B. In the more distal part of the branchlet, 

 fresh solenial outgrowths of the chief members give rise to new lateral members Avhich 

 end in anthocodiae towards the tip of the branchlet. The three chief members A, B, and 

 C can be traced to the extreme tip of the In-anchlet, where they emerge to form a terminal 

 group of three anthocodite. The larger lateral members (• and d emerge next below the 

 terminal group, and /next below these. Now, if the branch represented in section in 

 fig. 19 be traced downward into the stem, it is found that all the lateral members a, b, 

 ■c, d, e,fmonev or later make communication with the chief members A and B (in this 

 particular instance C does not give rise to lateral members in the branch and stem), 

 and end blindly below. A, B, and C can be traced down into the stem, where they form 

 lateral membei-s aud in turn make commiuiication with a chief member of the stem, thin 

 out, and end blindly Ijelow. Due or two of the most centrally placed chief members of 

 every branch are continued downward as chief members of the stem. From what 

 precedes, it follows that the growth of the chief members of the branching-system of the 

 anthodete is not arrested when they give off lateral members, but that they continue to 

 grow in advance of the lateral members derived from, and form the main axes of, the 

 branches and branchlets. Tiuis, as was explained in the introductory part of this paper, 

 the whole branching-system of any one primary zocid may be compared to a compound 

 raceme of which the lateral members run parallel one to another, aud are organically 

 fused together throughout the greater part of their com-ses. The whole anthodete is the 

 result of the organic union of several such branching-systems. But the racemose mode 

 of branching is not indefinite in L. iiitida. Some of the more superficial lateral members 

 of a branchlet are arrested in growth, but continue to branch on the system of a mono- 

 chasial cyme. Thus, the anthostele a in fig. 20 ends in au authocodia at the level of the 

 section, and does not grow further as an individual member. But as the sections are 

 lollowed iipward it is seen that the anthostele sends up a solenium lying in the angle 

 formed by A and c, and this solenium, some 2 mm. higher up, gives rise to the buds 

 represented under a liigh magnification in fig. 21. The anthostele marked x has exactly 

 the same relations. We find therefore in Lemnal'ia nitida a feature common in many 

 plants : namely, that the ultimate lateral members of a composite branching-system adopt 

 a mode of branching different from that of the chief member and its primary branches. 



It is to be noticed that in L. nitida, and the case appears to be similar in L. termiitalis, 

 there is not one, but three or four chief members in each branchlet. In Lemnalia 

 Jukesii, rhahdota, aud peristyla, the arrangement is simpler, for there is only one chief 

 member in each branchlet, and this emerges at the tip as the terminal authocodia. 



