THE CEYLON SPECIES OF CAULERPA. 95 



Here it may happen either that assimilators with a different number of branchlets shoot up from 

 the same horizontal axis, or that one and the same assimilator may have branchlets with a number of 

 rows varying at the top and at the base. 



For a discussion of these different variations and their nature it is perhaps advisable to begin by 

 describing some special cases. Let us, therefore, examine some Ceylon Caulerpas which show a varia- 

 tion of this kind. 



Among other forms of C. uvifera from the islands around Jaffna I observed one which I have des- 

 cribed under the name /. planiuscula (fig. 16). It shows a series of transitions in its branchlets. At the 

 base they are all ball-shaped, without exception, but a little higher up they begin to grow more and 

 more flattened, at first hardly noticeable, afterwards very evident. At the top of the assimilators tlie 

 greater number of the branclilets are more or less flattened, whereas at the base no such branchlets at all 

 are to be seen. 



An analogous variation of the branchlets is to be observed in C. corynephora f. complanata (fig. 18). 

 The main axis is flat, with the branchlets in two rows. At the base these are of the same width, but 

 higher up they swell out at the points, so that they have the appearance of being somewhat constricted. 

 The transition between the different kinds of branchlets is gradual, without sharp contrasts. 



To some extent of the same kind is the variation in C. dichotoma (figs. 23, 24), in that in this species 

 too the basal branchlets are simple and the dichotomous ones only begin higher up, but in tliis species 

 we have, besides this, to note that the branchlets at the base are arranged radially around the axis, where- 

 as the bifurcated branchlets higher up are arranged predominantly in two opposite rows, whereby the 

 whole assimilator becomes bilateral. Moreover it sometimes happens that some assimilators have only 

 simple radially arranged branclilets. C. dicJiotoma thus shows an example of different kinds of variation 

 of branchlets. 



C. peltata (figs. 31-33), on the other hand, shows somewhat different kinds of variation ; either 

 an exclusive increase in the assimilation discs, the intervals being then longer, at the same time as the 

 branchlets only develop from the upper side of the side branches, which thus become dorsi- ventral 

 (fig. 32), or some branchlets develop in the direction of the spherical form, showing a tendency to swell 

 up (fig. 33). The larger assimilation discs in the former variety, as well as the swollen ones in the 

 latter, occur at the tops of the shoots, whereas the base is surrounded by the more numerous but smaller 

 assimilation branchlets. 



Of about the same kind is the metamorphosis of the branchlets or pinnules in C. Chemnitzia 

 (fig. 26-30). At the base they are more or less pronouncedly cylindrical (fig. 27), of about the same shape 

 as in C. IceteiHrens ; higlier up they become trumpet-shaped at the top and as if abruptly cut short. 

 At the very top the branchlets are either disciform in certain varieties (fig. 29), which have a tendency in 

 the peltata direction, or spherically swollen in others which thus approach tlie racemosa-uvif era series 

 (fig. 30). 



In all the forms of variation in the branchlets described above it is a common feature that the 

 variation proceeds more or less regularly, step by step, and furthermore that a tendency to surface 

 increase is apparent, though it is effected indifferent ways ; for it is evidently shown in C. corynephora f. 

 complaruita (fig. 18), dichotoma (figs. 23, 24), peltata (fig. 32), Chemnitzia (fig. 27), and also to some extent 

 in uvifera f. planiuscula (fig. 16). 



Reinke, in his treatment of the variations of Caulerpa {loc. cit. p. 87), has distinguished between 

 such as pass over " allmahlich " and such as pass over " sprungweise," i.e., " graduafly " or by " leaps 

 and bounds." The variations in question are apparently identical with those REI^fKE has characterized 

 as " allmahlich." 



It is indisputable that this metamorphosis of the branchlets reminds one in many respects of the 

 heterophylly of tlie higher plants. It can, however, be of an essentially different kind, in that the difference 

 between the younger and older leaves is bound up with different conditions of life during different phases 



