THE CEYLON SPECIES OF CAULERPA. 91 



to the growth of clavifera, whereas such formations are rare in the north. There the shores are more 

 sandy, with very scanty vegetation in the upper parts of the httoral region, and with algal life beginning only 

 a little further down. That C. uvifera is closely allied to the sand Cauleipas, I have already pointed out. 

 Though it is chiefly the rhizonies and the root system in which this appears, yet even the assimilation axes 

 point to its being predominantly a deep form. For the assimilation axes are much longer than in C. davi- 

 fcra, and especially very thickly covered with branchlets, which show a firm cluster-shaped formation 

 (fig. 15). That such an organization is very unsuited to exist amongst corals and Actinians has already 

 been pointed out above ; it would be equally unsuited to strongly exposed locahties. Deeper down in 

 sandy bottom, however, it has a more suitable habitat, and the long vertical axes wave to and fro in the 

 currents so that the whole axis gets exposed to the light from every side. As JoHS. Schmidt {loc. cit., 

 IV., p. 105) also only found uvifera at a depth of from "1-2 fathoms water (coral sand)," the opinion I 

 formed about this plant in Ceylon is strengthened : the opinion, namely, that it is a sand Caulerpa from 

 somewhat deeper regions. 



With respect to the variations of C. uvifera and their ecology I need not take up much space, 

 since I have only collected /. planiuscula myself. It difiiers only in that its upper branches show a tend- 

 ency to flattening out (fig. 16), while the vertical axes at the base have rather few side branchlets. This 

 change seems to be favourable to the reception of light coming directly from above, but about the 

 special habitats of this plant my notes unfortunately do not afford me any help. It occurs only in 

 the north of Ceylon, with about the same distribution as the main species. 



C. longistipitata (fig. 45) and C. sedoides f. crassicaulis (fig. 47) have a similar organization to that 

 of C. uvifera. As far as C. sedoides is concerned, at least, the resemblance seems to apply also to its mode 

 of growth, for at Weligama it grew inside the reef at a depth at low-water of about 2 metres together with 

 C. taxifolia and sertularioides , anyhow not in the upper exposed part of the littoral region. At Matara it 

 occurred in rock pools. It also seems capable of variation in the length of its shoot axis, some branches 

 being very short (fig. 47), analogous with those in G. clavifera. 



C. longistipitata, on the other hand, seems to show some analogies with C. Icetevirens in its lengthy 

 root system that creeps along the stones, and its long, slender, and weak assimilation branches. It occurs 

 at Paumben in the upper littoral region, but not in any exposed places. Seeing that I only observed it in 

 that place, I will not venture to express any opinion as to where this form is reaUy at home, but JoHS. 

 Schmidt {loc. cit., IV., p. 105) also informs us that he only found it in the littoral region. 



Of the remaining species that occur in the littoral region, C. peltata, nummularia, parvula, imbri- 

 cata, and also Chemnitzia majy be considered as a very uniform and closely related group, distinguished by 

 aU their short branchlets being more or less pronouncedly disciformly flattened. 



Of these species C. nummularia (figs. 35, 36) occurs in about the same localities as C. clavifera, i.e., 

 in the upper littoral region, not rarely together with corals. The organization of the shoots is also quite 

 analogous with that of C. clavifera. The branches are aU recumbent, root-taking, and creep among 

 corals and on stones. These branches are dorsi-ventral and form disciform assimilation branchlets on the 

 upper side only. The whole plant resembles a matty tuft (fig. 36) with the assimilation system in one 

 level. Thus we here find an organization quite analogous with the one we have already described in 

 C. clavifera (fig. 13), and which is characterized by the relatively strongly developed root system in 

 comparison with the shoot system, and by the shortening of the shoot axis, whereby the assimilation 

 system is flattened out so that it receives the whole light vertically from above. C. clavifera and C. num- 

 mularia are both pronounced fight algae from the upper littoral region, growing in locahties with frequent 

 changes of water and often somewhat exposed to the swell. These analogies in the outward conditions of 

 life correspond to pronounced resemblances in mode of growth. Therefore, too, these forms might with 

 reason be characterized as ecological adaptations ; they are, as Detto says ("The orieder direkten Anpas- 

 sung," p. 146) examples of ecological convergence. 



