COSMAKIUM. 43 



We recorded this variety in 'Ale*. S. England/ 1897, p. 491, 



f CJ c_ x s J_ 



as "a small and rather irregular form." The forms seen from 

 the New Forest did not correspond exactly with Raciborski's 

 Polish ones, the angles of the semicells being somewhat more 

 rounded,, and the cells, as a whole, showed slight irregularities 

 of form. Schmidle has placed this variety under C. bicuneatn >/i 

 (Gay) Nordst., but this change we do not for the present 

 accept. C. bicuneatum may be distinct from C. arctoum by 

 reason of its great compression when seen from the side or 

 vertical views, but this apparent distinction demands further 

 inquiry. AVe have never seen any forms compressed in this 

 manner, and the elliptic vertical view of var. tatricum, com- 

 bined with its general form, indicates a very close relation- 

 ship with C. arctoum. 



73. Cosmarium decedens (Eeinscli) Racib. 

 (PL LXIX, figs. 6-8.) 



Cosmarium plicatum Reinsch ? C. decedens Reinsch, Spec. Gen. Alg. 1867, 



p. 114, t. 22, f. 7-9. 

 Didymidium (Cosmarium') plicatum Eeinsch, Algenfl. Frank. 1867, p. 109, 



t. 7, f. 1 d [non a c]. 

 C. sinuosum Lund. var. decedens Kordst. Desm. arctoee, 1875, p. 38, t. 8, 



f. 41 ; Desm. Ital. 1876, p. 31 ; De Toni, Syll. Alg-. 1889, p. 894 ; West, 



Alg. W. Ireland, 1892, p. 142 ; Roy & Biss. Scott. Desni. 1894, p. 174 : 



Borg-e, SiissAv. Clilor. Archaiig. 1894, p. 22; West & G. S. West, 



Alga-fl. Yorks. 1900, p. 90. 



Euasti'iim decedens Roy in Scott. Xat. July, 1883, p. 38. 

 Cosma;-ium decedens (Reiiisch) Racib. Desni. Xowe, 1889, p. 88; Borg. 



Freslnv. Alg. Faeroes, 1901, p. 221, t. 7, f. 9. 

 C. decedens Racib. b. carpaticinn Racib. 1. c. p. 88, t. 5, f. 1. 

 C. decedens c. boreale Racib 1. c. p. 88. 



Cells of moderate size, about twice as long as broad, 

 moderately constricted, sinus slightly open ; semicells 

 snbrectangnlar, sides concave, basal angles slightly 

 protuberant, and generally closely opposed to those 

 of the other semicell, upper angles somewhat produced 

 and prominent, scarcely rounded, apex widely ret use. 

 Side-view of semicell ovate-elongate. Vertical view 

 very broadly elliptic (sub-circular), with slightly pro- 

 duced (faintly mammillate) poles. Cell- wall delicately 

 punctate. Chloroplast axile with one central pyrenoid. 



Zvgospore unknown. 



Lenth 40-53 ' breadth 20-30 {JL; breadth of 





