FRESH-WATER A L G uE F THE UNITED STATES. 67 



w 



groundwork of the stratum. They appear to be unattached to the soil, and each filament 

 may be about half an ineli in length ; but they ai'e eummonly found brukeu ott' at the inferior 

 eud, or the lower part deeays whilst the upper continues to grow. They are slightly curved, 

 in serpent-like fashion, never quite straight ; at first they are simple, but now and then emit 

 lateral branches, which issue at considerable angles and generally in pairs. When a filament 

 is about to branch, a rupture takes place in the side of the sheath, and the endoehrome issues 

 in two portions, one connected with the upper, the other with the lower half of the filament; 

 these form the nuclei or medullary portion of two new branches and become duly invested with 

 a membranous sheath, and gradually put on the aspect of the adult filament. The endo- 

 ehrome is granular, dark-brown, and annulated at short intervals, the transverse rings being 

 placed very close together in the youngest portions, and less closely in the older, where they 

 are distant from each other about twice the diameter of the column. Tliis annulated endo- 

 ehrome is interrupted at certain fi.xcd places, where an ellipsoidal cell is formed, separating 

 the endoehrome of the lower from that of the upper portions." Harvey. 



Reniarhs. — I have never seen either the genns or species, and therefore am 

 forced to copy the descriptions of both from liabeuhorst and Harvey. 



Family SIEOSIPHONACE^. 



Thallus ramosus, e eellulis pachydermaticis aut uni vel pluri seratis et in vagina ampla inclusis 

 formatus, interdum eellulis perdurantibus instructus. RaniiUcatio vera fit cellularum vcgetativaruni 

 quaruudam divisinne in a.xis longitudinalis din-ctlonera, (|ua ex re cellula; dme sororias gignuntur; 

 cellula inferior in trichomatis continuitate permauet, superior divisione continua rcpetita in eaudcm 

 directionem se ad ramum e.xplicat. 



Propagatio adhuc ignota. 



Frond branched, formed of thick-walled cells in an ample shenth, sometimes furnished with liete- 

 rocysts. Cells uni- or multi-seriate. Branches formed by a longitudinal division of certain 

 cells, so as to form two sister cells; the inferior of which remains as part of the trichoma, whilst 

 the other, by repeated divisions, grows into a branch. 



Propagation not known. 



Remarks. — The Sirosiplionaceos are the most complex in tlieir organization of 

 all the Pliijf-ocliromophuceai, in so far as the protoplasm within the slicaths is every- 

 where broken up into a number of distinct cells, each of which is provided with 

 a thick coat or wall as well as in the circumstance of the frond having more perfect 

 branching. The so-called pseudo-branches in tlie other families are more truly 

 comparable to distinct fronds or thalli remaining attach(?d to the parent thallus 

 than to distinct branches, whilst among the sirosiphons the branches really belong 

 to the original thallus. The heterocysts are much more frequently absent than 

 present, only one of the known American species being furnished with them. 

 The sheaths are generally not so distinctly sheaths as among the oscillatoria, &c., 

 for, instead of being distinct tubes, they appear rather in most cases as masses of 

 firm jelly, the outer portion of which is hardened almost into a periderm, and in 

 the inner part of which the cells are imbedded. Their color varies from the 

 transparent colorlessness of glass to a dark opaque-brown. Their surface is per- 

 haps most frequently smooth, but at times is tubercidate or otherwise roughened. 

 I have never seen anything like spores about them. 



