THE MARINE ALGJE OF NEW ENGLAND. 131 



tripartite, clustered in involucrate whorls at the nodes or on the inner 

 side of short fascicled branches ; cystocarps (favellse) involucrate. 



A beautiful genus, comprising between 30 and 40 species, but only represented on 

 our Eastern coast by a single species aud on the Western coast by two doubtfully 

 determined species. The genus is distinguished from CaUUhamnion by the involucrate 

 favella? and by the disposition of the tetraspores. As we have Spermothamnion sepa- 

 rated from CaUUhamnion in consequence of the absence of the gelatinous envelope 

 found in true favelhe, so we have Bornetia separated in a similar way from Grijfithsia. 

 The genus can generally be recognized at sight by the rather large but very delicate 

 cylindrical, oval, or, at times, globose cells, which do not bear immersion in fresh 

 water even for a short time, and by the branching, which is dichotomous or a modifi- 

 cation of the dichotomous type. The accurate specific determination from sterile speci- 

 mens alone is generally impossible, so great is the resemblance of the fronds in the 

 different species. The antberidia vary very much in the different species. In our 

 only species they are sessile on the upper half of the globose terminal cells ; in G. coral- 

 Una they surround the nodes in tufts ; and in G. setacea they are in dense approximate 

 whorls, attached to the inner side of incurved branchlets. The tetraspores also vary 

 in the different species. In G. Borneiiana and G. coralUna they are in whorls at the 

 nodes, and are attached to the inner side of short simple branches, which form a whorl 

 around the node. In G. setacea the tetraspores occupy a position which corresponds to 

 tbat of the autheridia. The favellae are always truly involucrate and, as far as is 

 knowu, terminal, in our species occupying the place of a suppressed dichotomy. The 

 development of the procarp of C. coralUna has been fully studied by Janczewski. In 

 that species he found two triehogyues to each carpogenic system, as is also the case in 

 the genus Cerammm. A non-sexual mode of propagation, by means of cells which 

 give off root-like processes, has been described by Janczewski in G. coralUna, and a 

 similar process takes place in G. Bornetiana. 



G. Bornetiana, Farlow. (G. coralUna f Harv., Ner. Am. Bor., Part 

 II, p. 228, non Agardh.— G. globulifera, Kiitz., Tab. Phyc, Vol. XII, 

 pi. 30. — G. globif&ra, J. Ag. in part,— G. Bornetiana, Proc. Am. Acad., 

 1877.) 



Exs. — Alg. Am. Bor., Farlow, Anderson & Eaton, No. 88. 



Fronds dioecious and dimorphous. 



Male plant. — Globosely tufted, one to three inches high; filaments 

 repeatedly dichotomous; lower cells cylindrieal-obovoid, several times 

 longer than broad, becoming shorter and broader above ; terminal cells 

 globose-pyriform ; autheridia sessile, densely covering the upper half of 

 terminal cell. PI. X, Fig. 4. 



Female plant. — Two to five inches high, loosely tufted, filaments re- 

 peatedly dichotomous ; lower cells cylindrical- obovoid, becoming broadly 

 pyriform above and then gradually diminishing ill size toward the tip ; 

 favellae solitary on the upper part of the superior cells ; cells of involucre 

 10-20, unicellular, club-shaped, somewhat incurved. PI. XI, Fig. 3. 



Tetrasporic plant. — More slender than the female plant; tetra- 

 spores tripartite, densely clustered around the nodes of special branches; 

 cells of involucre short and suberect. PI. X, Fig. 5. 



On wharves, sponges, shells, and occasionally on Zostera. 



