sovie Uecent Publications concerning Diatoms. 31 



Diatom.V Ann. & Mag. Nat. Hist. 1901, vol. vii. pi. iv. 

 fig. 21. 



Achnantlies brevipes and A. subsessilis (p. 43). 



All true Achnanthidia have four plates united into pairs 

 by a common pyrenoid. This is also correctly represented 

 in Karsten's figures of both species, and in the description 

 given in the general part (p. 161) he mentions four plates ; 

 but when on p. 42 he says " Zwei Chromatophoren beiderseits 

 des centralen Kern," this can easily be misunderstood. We 

 differ, however, in one respect ; according to Karsten the 

 two opposite plates are united by a transverse band of endo- 

 chrome, while in my figures (Et. sur I'Endochrome des Diat. 

 pi. vi. figs. 20, 22) no such band exists, the two plates being 

 united by a colourless body. I have been criticized by 

 Karsten for calling it (as well as similar bodies in Okedenia 

 inflexa and 0. pontica) a pyrenoid. Indeed, a colourless 

 pyrenoid seems to be a contradiction in itself, as, accordino- 

 to the generally accepted opinion, a pyrenoid is a body colour- 

 less by itself, but being enclosed in the very substance of the 

 endochrome has therefore necessarily a coloured appearance. 

 This, however, is not quite correct. The pyrenoid can emero-e 

 from the chromatophore either partly {Mastogloia, Achnan- 

 thidium) or even entirely. We find all three cases occurring 

 sometimes in the same genus : coloured common pyrenoids 

 [Achianthidiufn [Cymbosira) Agardliii {K\X{z,y]^ partly out 

 of the endochrome {A. sp.), and completely free {A. glahra- 

 tum). To this latter type also A. hrevipes^ at least its var. 

 intermedia, seems to belong. I intend to return to this inter- 

 esting subject in a special note. 



Navicula dicephala, W. Sm. (p. 51), 



cannot be this species, as N. dicephala has the same endo- 

 chrome as Cymhella (one plate), and therefore does not belono- 

 even to the family Naviculaceas. 



Navicula mollis (W. S.), CI. (p. 5b, fig. 51). 



The third figure (next to the margin) is most probably a 

 Lihellus. 



Navicula humerosa, Breb. (p. 56) . 



Fig. 55 represents either a stage of division or else it is 

 an error, this species having normally two plates alono- the 

 valves (see Et. s. I'End. d. Diat. pi. ii. fig. 6) ; this is exactly 

 reproduced in fig. 56, representing, according to Cleve, a form 

 of ]S. humerosa. In a letter Karsten expresses his opinion 

 that the diatoms which we have both studied, altliougli very 



