BY G. I. PLAYFAIR. 121 



sions are: Cell, inatric. long. 50-65, lat. 39-52, autosp. long. 39, 

 lat. 19/x. The autospores, however, are wrongly delineated. As 

 drawn, they work out at 61 x 19 instead of 39 x 19. Turner 

 says: '^spcyi'is .... loitge ovaUhns,'^ bnt 

 his drawing them cylindrical makes it 

 certain that tliey were really oblong. 

 The ratio of length to breadth also (at 

 39 X 19) is just 2:1, as in O. NacieJll. 

 W. ct G. S. West, Frw. Alg. Madagascai; Text-fit/ 1 1 



p.82, identify this form (but doubtfully) Oocysfis elUpfk-a i. mivor 

 with O. eUijJtica West {= O. Ndyelii). W. We.st, ( x 520) ; after 

 The shape of the mother-cell is of no im- ^^'- & <^- S. West, 

 portance, cf. W. cfe G. S. West, /.o.,P1.5,f. 13-14 (Text-fig. nostr 11). 

 Nordstedt's Sandwich Island specimens of O. JVdgelii should 

 probably be placed under this variation. He gives long. 30-32, 

 lat. 14-15//, Alg. aq. dulc. Sandvich., p.8. I have not come 

 across this form in my gatherings yet. 



Var. OBESA, n.var. (PL vii., f.7). 

 Prte longitudine crassior. Cell. long. 20-25, lat. 13-16/i. 

 Guildford (146). 



Var. CURTA, n.var. (PI. vii., f.8). 



Late-ovalis vel brevissime oblonga. Cell. long. 17. lat. 13/x. 



Guildford (146) cum priori rarius. 



Syn., O. Ndgelii in Lemm., Plankt. Schwed. Gewass., p. 107, 

 where he remarks : " Zellen breit-oblong, 15-17 : 1 1-13/x." Both 

 forms quite common in No. 146, from a pond at Guildford; the 

 one is evidently a reduced form of the other. The ratio of length 

 to breadth in var. obesa is from 1 "4 to 1 "7 : 1 '0 as compared with 

 2:1 in the type. In var. curta, it is 1 -3 to 1 0. Chloroplasts 

 not noted, the contents deranged. 



OocYSTis NoViE SEMLiiE Wille. (PI. vii., f.9, 10). 

 "O. cellulis homogeneis('?) singulis 1. in familias e 4-8 cellulis 

 formatas consociatis, familiis nonnunquam in massa gelatinosa 2-4 

 consociatis; membrana crassa, non tuberculo apicali instructa." 

 Wille, I.e. /v - 



LIBRARY 



