FEESHWATEE ALG^. 27 



Var. ANTARCTICA var. nov. 

 (PI. II., figs. 87-90.) 



Fills eloHgatis, rectls vel flexuosls, Interdum tortuosis ; vaglnls hyallnls, plus 

 minusve tenuibus, levibus, ut videtur, setate provecta non erassis ; trichomatibus 

 serugineis, apice baud vel raro attenuate, truncate, ssepe massa intercellular! inflate 

 et refringenti interruptis ; cellulis diametro trichomatis quadruple ad sextuple 

 brevieribus, cyteplasmate ssepe cum granulis magnis, dissepimentis plerumque non 

 granulatis ; apice tricbematis interdum capitate, membrana cellulse apicalis superne 

 plus minusve distincte incrassata in ferma calyptrse. 



Diam. fil. = 8-13 ju.; diam. trich. = 7-12 /a. 



Hab. — Same as tbe type, but the variety was considerably more abundant. 



At first there seemed to be good grounds for establishing this variety as a 

 distinct .species, as it difi'ers from the type in several prominent respects, viz., 

 the thin character of the sheath, even in older filaments ; the rare attenuation of 

 the apex of the trichome ; the frequent presence of a prominent calyptra and 

 the marked capitate habit of many of the trichomes ; the presence of numerous 

 biconcave masses of refringent intercellular substance, which generally bulge a little. 

 None of these characters except the thinness of the sheath and the occurrence of 

 intercellular masses is, . however, constant. The apex of the trichome may be flat 

 and rounded without any development of a calyptra or even a perceptible thickening 

 of the outer membrane (fig. 88) ; or the apical cell may have a prominent calyptra 

 without possessing a capitate character (fig. 89) ; or the apex may be distinctly 

 capitate with only a slight thickening of the outer membrane (fig. 90) ; or lastly 

 the apex -of the trichome may be more or less markedly attenuated, capitate and 

 provided with a: well-marked calyptra (fig 87). It is obvious that these different 

 forms .of apex connect the variety with the type, a connection which is also 

 established by the identical width of the trichomes in the two cases and the 

 extremely flat cells. Should the diff"erent types of apex be constant we would 

 have £ attenuaia, f. capitata, f. simplex, etc., of this variety, but it is possible that 

 some or all of them are merely stages in development. 



The apical cell in the capitate trichomes was always very transparent and devoid 

 of contents. Occasionally th5 two end -cells of the trichome showed these characters. 

 In a few cases the cells near the apex of the trichome exhibited a faint constriction 

 at the dissepiments (fig. 89). 



32. Lyngbya martensiana. 



Lynghja martensiana Menegh., Gonspect. Algol, euganese, 18.37, p. 12 ; Gomonfc, op. cit., xvi. (1892), 

 p. U.5, PL III., fig. 17. 



Diam. fil. = 8-9 jlc ; diam. trich. = 6-7 " 5 /a. 



Hab. — Freshwater pond in ice off" Black Island, McMurdo Strait, December 

 31st, 1902; Granite Harbour, Januarj^ 20th, 1902, 



