302 Transactions. 



ment. Various modifications of and variations from the main types have, 

 been enumerated by Lang, and others also, in this paper, which appear 

 to supply connecting-links between them. The present writer would 

 emphasize especially the variations in the prothallus of L. ramulosum 

 in this connection. In respect of the character of the embryo types 

 also, Bower and Lang both contemplate a certain degree of affinity 

 between the different parts of the genus, with this difference in their 

 views : that the latter would look upon the L. cernuum type of embryo, 

 with its protocorm, as primitive or the genus, the i^rotocorm having 

 been lost from the life-history of the species which belong to the other 

 sections, whereas the former regards the L. Selago type of embryo as 

 being the most primitive, there being derived from it the type which 

 shows the large foot, while the L. cerntium type takes a middle position. 

 In the present paper the conclusion has been arrived at from the study of 

 the protocorm of L. laterale and L. ramulosum, and also of L. cernuum, 

 that this organ, as Bower suggests, is a physiological specialization rather 

 than a highly primitive feature. The comparative study of the vascular 

 cylinder in the genus Lycopodium also suggests that the different sections 

 as dealt with above are more or less closely related. Jones (13, pp. 27—28) 

 believes that a simple radial, root-like structure of the vascular cylinder is 

 to be regarded as the primitive one for the genus, and my own view, 

 derived from the study of the above eleven New Zealand species, is the 

 same. The growth in size of this type of cylinder would either leave the 

 configuration strictly radial or (on account of the constant changes in the 

 disposition of the individual vascular elements and of the groups and 

 rows in which they are arranged, a feature which is a characteristic one 

 right through the genus) would cause it to adopt a stellate form in which 

 cross-connections are present. Now, this is the type characteristic of 

 the Selago and Phlef/maria sections. From this the mixed type of the 

 Cernua and Inundata sections on the one hand, and the parallel type 

 of the Clavata section on the other hand, would be derived, as has 

 been described in Section V of this paper. Lastly, Miss Sykes (15) 

 has shown that Pritzel's sections of the genus can be arranged in a 

 series in respect of the morphology of the sporangium-bearing organs. 

 From her study, however, she would conclude that the Cernua section 

 comprises the most primitive members of the genus, and that those of the 

 Selago section have been derived from them by reduction, while the type 

 of L. inundatujn is " an interesting intermediate step." 



Thus a belief in the interrelationship of the different sections of 

 the genus Lycopodium is more generally held than the opposite one — 

 that they have been widely separated from an ancient period. The 

 view which the present study seems to suggest is that on a general sum 

 of characters the Selago section must be held to comprise the most 

 primitive members of the genus, and that the Phlegmarm and Clavata 

 sections have been independently derived from it, the former being less 

 modified than the latter. The exact relation of the Cernua and Inun- 

 data sections to the Selago section is difficult to gauge. The consensus of 

 opinion seems to be that the L. cernuum type of prothallus is to be 

 regarded as primitive for the genus, but the plagiotropic habit, the mixed 

 type of vascular structure, and the protocorm, on the other hand, would 

 all seem to be highly specialized characters. Hence the Cernua and 

 Inundata sections may perhaps best be jDlaced in a group by themselves, 

 as having been derived from ancestors common to themselves and to the 

 Selago section, but independently of the latter. 



