RHODOSPERME.E. 73 



dently metamorphosed ramelli. But how shall we accomit for 

 the walls of the ceratnidiiim? We must either suppose these 

 walls to be formed of the union of the first developed or low- 

 est ramelli; or, w^hich I think more probable, consider the ce- 

 ramidium as an introverted branch, whose apex either turned 

 inwards, or stopped short while the cells of the periphery con- 

 tinued to grow around and above it, until they finally enclosed 

 it. In either case the ceramidium would be formed as shown 

 in the imaginary diagram annexed. The cells of the walls are 

 always of irregular shape and irregularly placed, as if pushed 

 from their position. Such dislocation would naturally result 

 from the continued lateral growth of the radiating cells of 

 the apex of the branch, after the cessation of growth up- 

 ward, however that cessation originate. Were the walls 

 formed from the coherence of ramelli into a membrane, we 

 should expect to find their fibrous origin indicated in the 

 structure of the membrane ; but no fibrous structure appears 

 in the walls. Whatever the nature of the walls of the cera- 

 midiuni, there seems no reason to doubt that its contents — 

 the tuft of spore-bearing fibres — are a metamorphosed condi- 

 tion of the ramelli, which would have tipped the cerami- 

 dium, had it been developed into an ordinary branch. I 

 have already endeavoured to show that the ramelli are 

 the representatives of leaves ; and if the structure now attri- 

 buted to the ceramidium be considered established, we shall 

 have strong analogical evidence in favour of the spores being 

 considered of the nature of seedn, and not as buds ; and by 

 consequence, that tetraspores should be regarded as gem- 

 mules. For we find, in flowering plants, transformations 

 strikingly similar to what I have been describing. In them 

 the flower is a truncated branch, and all its parts are meta- 

 morphoses of leaves : this flower produces seeds. In Poly- 

 siphonia, &c., the ceramidium is a truncated branch, and its 

 contents are modifications of ramelli (or supposed leaves) : 

 this ceramidium produces spores. The seeds in the first 

 case, and the spores in the second, are formed — so far as po- 

 sition goes — under analogous circumstances ; and therefore 

 it seems not unreasonable to infer that an analogical rela- 

 tionship exists between them. More than this I do not 

 consider established, for we do not yet know the cause of the 

 formation of conceptacles and the production of spores. We 

 know that seeds result fi-om the joint agency of stamens and 

 pistils. But we do not know whether any process similar to 

 fertilization takes place with the spores of these Algae. It 



