8 



AMERICAN HYDBOIDS. 



first an ordinary monosiphonic stem, and that the canaliculations are formed by portions of this 

 common c<enosare becoming surrounded by tubes of eliitin. Leaving the center of the stem 

 entirely vacant, the canals being peripheral. I am not satisfied, however, that each of these 

 canals is surrounded by perisarc. 



Tin Jtr<i Hi-lux. - There is but little that is characteristic to be found in the branches of the 

 Sertularidse. Their disposition has already been discussed. They never become highlv special- 

 ized to form protective structures for the passive or active defense of the gonophores, as they so 

 often do in the Plumularida 1 , where, they form the curious phylaetocarps. They do, however, 

 become modified to form accessory tubes in the fascicled stem, as we have seen, and they often 



become tendril-like at their distal ends and clasp 

 other branches so as to form anastomoses and 

 sometimes a flabellate structure, as in the genus 

 Dictyocladvum. 



There appear to be two methods by which 

 the attachment of these new tendril-like branches 

 is formed. In I>. ilii'ltntiniiii. "when a branch 

 is destined to form a union of this kind its distal 

 extremity becomes elongated into a tendril-like 

 continuation destitute of hydrothecse. When 

 this meets a neighboring branch, the end of the 

 tendril unites with a branch, not however with 

 any part of the surface of the branch indiffer- 

 ently, but, directing itself toward the orifice of a 

 hydrotheca, it here attaches itself, its axis be- 

 coming directly continuous with that of a hydro- 

 theca." 1 In the only other species of the genus 

 thus far discovered, D. flahellwn Nutting, the 

 connection is formed much more simply. The 

 tip of the tendril-like process grows around the 

 branch that it meets and clasps it very much as 



does the tendi'il of a vine. (Fig. 



The adhe- 



THE .CvASTiOInsIS OF nnANCHKS IN HICTYi ii'LAI'Il'M. 



Fig. 7. Tip of branch uf liirt;ii><'/<nl/ii/it j/>iL, //,<//i clasping another 



branch. 

 Fig. 8. Another brunch termination which forms lulu-like ].ru- 



ci sses tn M-curr liriiier attachment to another branch. 



sion is quite strong and permanent, and the clasp- 

 ing portion of the tendril may become eventu- 

 ally embedded in the branch so as to be hard to 

 see, being indicated merely bv a swollen area 

 over the original point of contact. Sometimes 

 the growing tip upon meeting a branch spreads 

 out in lobular projections, thus increasing the 

 extent of the surface of adhesion, as in tig. 8. 

 This same method of joining different brandies 



is found quite commonly in several species of Xr/(/A//v//</, especially those of a straggling habit 



of growth, such as .S'. <l/rliuti>nitt and S'. , 1,-i/mix. 



A merely temporary joining of the tips of hydroeladia has been reported in the case of a 



species of AijluopJietuii." This, however, seems to be for an altogether different purpose from 



the one subserved in the examples described above, and served to join two colonies, apparently 



for the purpose of conjugation. 



I have seen no sertularians, aside from the genera X rtnliii'i 11 <i and 1 >i<'ti/<>i-ltif]i inn, in which 



anastomoses of branches occur, except perhaps as a rare abnormality. 



Tin: Ifi/ili'iintJi. I have been unable to find any evidence of the hydranth being studied with 



care in any sertularian, at least since modern histological technic became available. Without 



this aid. however, Louis Agassiz was able to make out most of the essential points, even of histo- 



'AlhiKin, t'/i, i//, i,,/, ,- llepurt, Hydroida, Second Part, 1885, p. 77. 



2 See Part I, p. 45. 



