V. SPOXGIOCARPE.E. 127 



composed of slender, dichotomous filaments, exactly like those of the cortical layer 

 from which they spring. Among these filaments, within the wart, are formed 

 innumerable globose nuclei or spore-clusters, attached to the filaments ; each sur- 

 rounded by a broad pellucid limbus and consisting of many obconical spores 

 radiating from a minute central placenta. Each spoi-e is enclosed in a gelatinous 

 perispore ; and the pellucid limbus which surrounds the nucleus seems to be formed 

 by the contact of the numerous gelatinous perispores. It is therefore to be 

 regarded as spurious. 



Tetraspores are formed in the upper, slightly swollen branches. They are 

 deeply sunk among the peripheric filaments, oblong, and at length cruciately 

 parted. 



This Order is founded on a single genus composed of a solitary' species, so 

 remarkably distinguished by its fructification from all other known genera of 

 Algte that it can scarcely be referred without violence to any established group. 

 In external habit of the frond, as well as in internal structure, there is a very close 

 resemblance between Pohjides and Furcellaria (a genus of Cryptonemiacecv) , so close 

 indeed that these two genera have frequently been united, and even in the last 

 work of Professor Kiitzing, an author not remarkable for avoiding generic analysis, 

 Polyides rotundus is described as a species of Furcellaria. In the fructification of 

 these genera, however, there is so wide a difference that if we are to regard struc- 

 ture of the sporiferous nucleus as a surer guide to natural affinities than structure 

 of frond, we must place them at nearly opposite ends of the systematic arrange- 

 ment. Another and very opposite affinity f )r Pohjides has recently been indicated 

 by Professor J. Agardh, who, in his tabular distribution of the Floridea?, places it 

 among his Chondrieoi, the proper type of which group is Lanrencia. This relation- 

 ship is no doubt inferred from the size and form of the spores, which may be com- 

 pared with those oi Lomentaria rather than with those of any other of the Chondriece ; 

 but in the absence of any indication of nearer affinity between these genera, the 

 mere form of the spores will hardly be thought sufficient. The appearance and 

 structure of the fronds are very dissimilar, and the details of the fructification 

 abundantly unlike. In the Chondriece (our LaurenciaceoB) a single sporiferous 

 nucleus is enclosed in a hollow conceptacle, formed out of the end of a truncated 

 branch, and of the most perfect tijpe wliich the conceptacle attains among the Rliodo- 

 spennatous Alga'. The contained nucleus is attached to a basal placenta, and 

 therefore terminates the axis of growth. Such a structure is extremely different 

 from what we have above described in Polyides, and though very unwilling to mul- 

 tiply Orders, I cannot consent to the ordinal association of genera differing so 

 widely in fructification. In my opinion Polyides is more nearly related either to 

 the Gelidiaceoe or Helminthocladiece, though scarcely to be placed among either. In 

 the external aspect of the fruit there is much resemblance to Peyssonnelia, but 

 much dissimilarity in the nature of the nucleus. So long ago as 1830 Dr. 

 Greville proposed Polyides as the type of a separate Order, and I now revert to 

 his views. 



