COSMAKIUM. 153 



isthmus 10-11 '5 A*; thickness 22-24 /t ; diam. zygosp. 

 without spines 46-50 /u, with spines 70-95 /u. 



EXGLAXD.- -Cumberland! Westmoreland! AY., X., 

 and E. Yorks ! Leicestershire (Roi/). Essex (Rulfx). 

 Surrey (with zygospores from Thursley Common) ! 

 Sussex (Rolfs). Hants (Rolfs) ; zygospores from Xew 

 Forest ! Devon ! Cornwall ! 



WALES. --Capel Curig (Cooke $ Wills), Bethescla, 

 Llyn Idwal, Llyn Ogwen, Llyn-y-cwm-ffynon, and 

 Moelfre, Carnarvonshire ! Llyn Coron, Anglesey ! 



SCOTLAND.- -General ! (Roy fy Bissetf). Rare in the 

 plankton ! Common in Lewis and Harris, Outer 

 Hebrides ! 



IEELAXD. Donegal! Mavo ! Galwav ! Kerry! 



O .' ,/ ii 



Dublin and Wicklow (Archer). Down ! Eare in the 

 lake-plankton of Galway and Kerry ! 



Geogr. Distribution.- -France. Germany. Austria 

 (including both Bohemia and Galicia). Poland. Italy. 



Portugal. Xorwav. Sweden. Bornholm. Denmark. 



-r t 



Finland. X. and S. Russia. Iceland. United States. 

 Brazil. Paraguay. 



O i 



Cosmarium ornatum is a widely distributed species, occur- 

 ring in bogs and at the margins of lakes. The widely truncate 

 apex of the semicell projects very slightly, and furnishes the 

 most constant and characteristic feature of the species. The 

 general granulation of the semicell is fairly constant, but the 

 granules covering the central protuberance are exceedingly 

 variable in number and disposition. They may have a con- 

 centric disposition, or they may be arranged in somewhat 

 irregular vertical series, and sometimes their disposition is 

 quite irregular. In all cases there is a small,, smooth area 

 surrounding the central granules and separating them from 

 the remaining granules of the semicell. 



ij C? 



The figures of this Desmid o-iven bv Ealfs in his ' British 



t/ 



Desmids/ 1848, t. 16, f. 7, are not very good. ~\Ve believe 

 the central granules to be incorrectly portrayed, and the 

 amount of projection of the apices is distinctly exaggerated. 

 These inaccuracies caused Lundell to state the differences 

 between his Swedish specimens and the ng'ures given by 

 Ralfs, and led to the institution of " var. anglica '"' and " var. 

 suecica' bv Eaciborski. We have examined numbers of 



