132 



CEYLON MARINE BI0L0C4ICAL REPORTS. 



increase in size. Moreover the side axes have a pronounced tendency to develop their disciform branchlets 

 on one side only, i.e.. on the upper side ; which apparently depends on their position in relation to the 

 horizontal level. For, horizontal axes form their assimilation system only on the upper side, vertical 

 ones all round, which is natural from the point of view of their exposure to the Ught of most advantage 

 to them ; this is especially the case if these horizontal axes take root. A transition to C. nummularia is 

 thus effected, since, as I will show later on, it is just this form that is characterized by its assimilation 

 discs being one-sidedly developed on the horizontal procumbent main axis. 



Also in another respect this form is remarkable in that, as fig. 33 shows, some of the disciform assim- 

 ilation branchlets are somewhat swollen and thus would very closely remind one of c?aw/era branches, if 

 a rather evident border did not give evidence of the branchlets having originally been flat (fig. 34). Weber 

 v. BossE mentions about C. macrodisca (Agardh), that its great branchlets are " distinctementbombes 

 dans lean pour devenir bientot plates quand ils en ont ete retires." While then in the case of C. macro, 

 disca it seems to be the rule, in peltata it is rather the exception, the majority of the branchlets having qiute 

 flat assimilation discs. It is to be noted that C. nummularia, too, seems to have the same tendency. So 

 Harvey's Friendly Island No. 76 (Herb. J. G. Agardh, 1681 4), very evidently shows a mixture of nummula- 

 ria and davifera branches. This form Harvey has called G. clavifera v. platijdisca. 



Fig. 34. — C. peltata (lam.) f. ad claviferam, 

 swollen branchlets from the upper 

 part of an axis. (3 X 1). 

 Fig. 33. — V. peltata (lam.) 1'. ad claviferani 

 (1 X 1). 



Oeographical distribution. — Ceylon : in the somewhat deeper parts of the littoral zone : Galle I 

 Kosgoda I Kangesanturai I Jaffna ! distribution for the rest uncertain, because of the incompleteness of 

 the notes previously made about it. 



I have seen unmistakable specimens of this form from the Red Sea and from the Java Sea 

 (Batavia). 



16.— CAULERPA NUMMULARIA (Harvey) Rbinke. 

 Rbinke, Ueber Calerpa, p. 39. 

 J. G. Agardh, Till. Algernes Syst. I., p. 38. 

 Exsicc. : Harvey, Friendly Island Algae, No. 77. 

 Peroitson, Ceylon Algse, No. 101. 



Under this name Harvey distributed a Caulerpa from the Friendly Islands (No. 77) which was 

 more exactly described by J. G. Agardh ("Till. Alg. Syst." I., p. 38). As characteristic of this form 

 he accentuated its great assimilation discs which are often crenulate at the border, and from these 

 crenules new stalks and discs grow out. But Agardh also points out that the species is closely 

 allied to /x'/toto and that the distinguishing character (i.e., the crenulate discs) is far from being constant, 

 for smooth-edged and crenulate assimilation discs occur together from the same horizontal axis. Weber 

 V. Bo.s.sE also accentuates this character as the main character for her var. nummularia. Reinke, however, 

 who in his treatment of the Caulerpas has a sharper eye for the morphopogical differences between the 

 types, finds the boundary between peltata (main form) and nummularia to consist in the mode of branching. 

 He says : Bei nummularia fehlen die Axen der Assimilatoren gewohnhch ganz, so dass die kreisrunden, 

 gestielten Bliitter direUtdein Riiizom cntspringon'" (" Ueber (;auler|)a," p. 39). With tins Reinke seems. 



