THE CEYLON SPECIES OF CAITLERPA. 



135 



Kjellman accentuated that the generally shortish vertical axes are thickly clothed or perfectly covered 

 from base to apex with branchlets, which are disciformly flattened at the apex with a diameter of 3-4 mm. 

 (fig. 39) Tliis character is quite sufficient for distinguishing this little Caulerpa, which in Ceylon is rather 

 common in the Uttoral zone. Figs. 37 and 38 show pictures of it. It is generally tufted very thickly 

 and seen from above one sees nothing of such a one but the close vertical axes nearly quite covered with 

 branchlets. 



The origin of this form can be imagined in various ways. One is that it is derived from a 

 parvula form (see below) in which the branches have been erected and have afterwards radiated, at the 

 same time as the branclilets have become somewhat larger. Some transition forms (/. minor) between these 

 two species — which have the mode of branching of C. imbricata and the size of the discs of C. parvula — also 

 point to the fact that such a development must not be considered as only a hypothesis and without 

 any real foundation. As a matter of fact, these forms also occur in almost the same localities in the 

 upper part of the littoral zone. 



But C. imbricata can be imagined also to have originated in another way. If we study its organi- 

 zation we might give this as its characterization : C. imbricata is a somewhat dwarfed C. uvifera in which 

 the branchlets have become somewhat disciformly flattened. In reality forms of this kind do occur which 

 derive directly from uvifera by change of spherical branchlets to flattened ones. C. uvifera f. planiuscula 

 is such a form. Moreover, there are some transition forms collected by Kjellman at Galle and by mjrself 

 at Jaffna, which point to a relationship with peltata main form (fig. 41). If in such a one the vertical 

 branches are shortened and get a more one-sided direction, at the same time as the discs grow smaller 

 and closer, there then arises an imbricata form. Hence it seems probable that such a form could be 

 derived from either the one or the other. 



Fig. 40. — C. imbricata (kjellm.) 

 /. minor n. f. (1 X 1). 



Fig. 41.— C. imbricata", Fig. 42.— C. imbricata{K3Ei.LU) 



(kjellm.) i. ad peltatam. (1 x 1). t.micctan.i. (1 x 1). 



It is rather remarkable that variations of C. imbricata are to be found which are quite analogous 

 with C. Chemnitzia. Fig. 42 shows such a form, for which I propose the name mixta. The basal branch- 

 lets are cylindical, about 3 mm. in^ length and only IJ mm. in breadth, and at the apex abruptly 

 cut off , as if truncated, just as in Chemnitzia. Somewhat below the middle of the vertical axis commence 

 the typical imbricata branchlets — with the discs sharply defined from the stalk — and continue up towards 

 the apex. This form, however, is not C. Chemnitzia, for the latter has not only basal branches which 

 are cylindrical, but its upper branches increase only slowly in breadth and are almost trumpet-shaped. 

 (^/. fig. 27), and often also the upper branchlets are spherically enlarged as in racemosa. Weber v. 

 BosSE had also placed Chemnitzia under this species, but with the observation that it is an intermediary 

 form between racemosa and peltata. In the form mixta the basal branchlets are cyUndrical as in 

 Chemnitzia, but the upper branchlets are typical imbricata branches. The mode of growth of this form 

 in the upper littoral zone corresponds also in every respect with that of C. imbricata. 



By reason of C. imbricata in Ceylon seeming to be a form or race equally distinguishable from the 

 remaining peltata forms, as are nummularia and typica (Weber v. Bosse). I think there is little 



