238 Siphonales 



filaments, which forms the main substance of the joint. The small cortical 

 branches, which are really only the terminations of the filaments of the 

 subcortical layer, are so compact that in surface view they have a more or 

 less hexagonal appearance. 



The axial filaments usually branch trichotomously, the middle branch continuing its 

 course as a portion of the central strand (E. S. Gepp, '04). After a time the filaments 

 cease growing and form the resting apex of the segment or joint. At this point they are 

 in close contact and communication is established between them in one of three ways : 

 (1) Free communication is established by means of large apertures in the walls of all the 

 filaments, so that the whole central strand becomes welded into a connected mass. The 

 individuality of the filaments is not lost, however, and on renewal of growth each filament 

 continues again its individual course as it branches out to contribute to the tissues of a 

 new joint (characteristic of H. macroloba and //. incrassata). (2) The filaments fuse 

 in groups of two or three, their separate identity being completely lost, and at the end of 

 the resting period the fused portions continue their growth as single filaments which by 

 the usual type of branching form a new joint. (Found in H. Tuna, H. gracilis, etc.) 

 (3) The filaments always fuse in pairs and their identity is not lost as they appear again 

 almost at once as separate filaments. This type is a distinguishing feature of //. Opuntia, 

 occurring in no other species. These three methods of communication between the central 

 filaments furnish useful taxonomic characters (E. S. Gepp, '01). 



Reproductive organs are only known in Halimeda. These are zoogoni- 

 rlangia producing large numbers of very small biciliated zoogonidia, and 

 they arise on clusters of dichotomous fructiferous filaments developed on the 

 dorsal margins of the segments. The fructiferous filaments, which are direct 

 continuations of the central filaments of the joint, grow out either in small 

 isolated tufts (H. gracilis) or as a continuous fringe along the dorsal margin 

 of the segment (H. Tuna). Each fructiferous filament arises only as the 

 result of the fusion of two of the filaments of the central strand. The 

 zoogonidangia are developed in clusters on short branches of the fructiferous 

 filaments, and are green in colour. The germination of the zoogonidia has 

 not yet been observed. 



An interesting point, the significance of which is not quite clear, is the fusion of 

 filaments of the central strand before any fresh growth takes place, whether vegetative or 

 reproductive. " Inasmuch as these fusions are found to precede the formation of new 

 joints on the one hand, and of sporangiophores on the other, the obvious inference is that 

 the fusions provide a powerful stimulus for further growth, whether vegetative or repro- 

 ductive. They form new and vigorous growing-points. But what the factor may be 

 which determines whether sporangia or whether a new joint shall result from the fusion is 

 veiled in mystery " (E. S. Gepp, '04). 



The interesting genus Penicillus (fig. 153) has probably arisen from 

 ancestors common with those of Tydemania and Halimeda. It has a cylin- 

 drical or slightly flattened stalk consisting of closely interwoven filaments 

 thickly incrusted with calcium carbonate. At the summit of the stalk the 



