93 



forms wliic'li are alike in struetiii'o only, wliile they, in other le- 

 spects, eaii lianliy he emisidered to belon;;- together. 



When thus (Jitrxow i'urther remarks tliat there is the most 

 complete ti'ansition between C. K!<tcro))ii>liali(s antl ('. oniltis iridis 

 anil further to C. iKdiatiix, antl wlien on tlie otiier hand lie h)oks upon 

 t', irutralis as a vai'iety of C. itxti'i-oniitlKdns. lie has surely «.^onc 

 too far. 



On the other haiul, there are thick walled, small, coarsely 

 sti-uctured foi-ms, which have quite a dittcrent appearance to the 

 ordinary C. rndidfun. and yet which must also be entered under 

 this species. Such forms are C. dcriun A. Schm. (L. 128) pi. (io, 

 f. 1—4 = C. radintus f. minor A. Schm. (L. 127, pi. 3, f. 34.) 



I have, stransre to say. never yet seen any really jrood drawing 

 of this characteristic species. As it is, however, - ;» indeed is 

 the case with all sjiecies of Cosc'moiViscm — very difficult to figure 

 properly, I must at present give up the thought of giving any 

 figure of it. 



Cleve's meaning with regard to C. niduttus is not quite clear 

 to me, as he (L. 40, p. 321) refers to Schmidt's atlas pi. 60, f. 9, 

 which does not appear to be at all a characteristic figure of C. 

 nidiatu.-', as this species is looked upon by Grax and myself. Cleve re- 

 fers too to Giuxuws remarks in Diatomeen from Franz Josefs Land 

 (L. 83, cfr. above) and mentions in Phytopiankton {L. 27, p. 23) 

 that C', radifttus is scarcely more than a little form of C. oculus 

 iridis. Cleve's opinion of the latter species does not, however, 

 coincide with Grunow's and Rattray's. For further particulars 

 see C'. siililiiiHit'iis. 



Rather common. More abundant in deep water samples than 

 on the surface. 



Distribution : Ratiier common all the year round on the 

 coasts of Norway, both the western and northern. Also widely 

 distributed on the European coasts of the Atlantic and its arms, 

 right lip to the arctic coasts (GJreenland, Jan ^[ayen. Spitzbergen 

 and Franz .losephs Laud.) 



C. centralis Eiiitu., Rattk. 

 (PI. VI, tig-. 1). 



Rattk. L. 124. p. 555. 



This species is very easily recognized, but has been confused with 

 C', onihci iridis and C. concinnus. It is sometimes not so easy to 

 distinguish it from the latter species, and it is possible that one 

 will not be able with certainty to keep them separate; but the 

 ditt'erence from C. ocnlus iridis in {Grunow's meaning) as Avell as 

 from C', radiatus is vci-y considerable. 



The characteristic marks of the species are tiie following: 



Valve considerably convex (pi. VI, f. 1). Structure radiate with j 

 dichotomously branched radii and rather coarse structure (though 

 finer than that of C', radiutus). The markings from a rather large 

 central rosette of even size to "A radius, then gradually somewhat 

 smaller (not suddenly small near the bordei'). Near the margin j 

 numerous fine apiculi (in a single row) and with about 120°s space 

 between them. F.esides two much larger, short linear, unsymmetrical, 

 margmal ones. 



Varies considerably in delicacy of structure and so often clo- 

 sely resembles C', conciimus, in common with whicli species it has the 

 2 characteristic unsymmetrical marginal and the numerous small 

 «ubmarginal apiculi. These latter may be indistinct, and sometimes | 



(hut rarely) not t<t lie found at all. es|ieeially when the structure 

 is rather coarse. 



(^i,Kvi;, who also occasionally mentions C. n-ntralis na occurring 

 in the jdanktoii, seems to regard this sjiecies as a less distinct form, 

 as a transition between -f. rourinmis and C. ocnlus itidis. As a 

 type for the latter species he quotes (L. 40, p. 319) C'. asU-rom- 

 plutliis riir. Iii/lridu iinvs. Franz .lo.sephs Land (L. 83) t. 3, f. '.). 

 This may perhaps show that Ci.eve considers tho.se forms of V. 

 centralis which are without distinct marginal apiculi and are of a 

 coarser structure to lie C. orulis iridis. 



In the important work before mentioned on the distribution of 

 species of plankton in the Atlantic (L. 401, the name C'. mitniUs is 

 omitteil, the coarser forms probably being reckoned as C', oculus iridis, 

 and the finer, with marginal apiculi. as C', concinnus. 



With respect to difference from C', concinnus, reference is made 

 to this species, where the structure is more particularly mentioned. 

 Moreover, there appears as a rule to be great diff'erences between 

 the two species in the living plankton; C'. cojicniMTi.v developes cells, 

 which are high with very thin walls and consequently very ea.sily 

 altered in shape, and with a strongly convex marginal zone, while 

 C. centralis has rather low, thickwalled, firm cells, more flatly ascen- 

 ding towards the centre. 



As before mentioned, it sometimes - but only .seldom, judging 

 from my experience — seems to be difficult to discern between 

 C', concinnus and C. centralis, and Grax (L. 70, p. 167) seems 

 to suppose that the difference depends upon variation hi salinity 

 and temperature. My impression is, that they are two comparati- 

 vely young species, but that they have already sufficiently distin- 

 guishing characteristics to enable them to be dealt with a.s specific- 

 ally ditt'erent forms. 



This species and the C', radiutus are the most freriuent of this 

 genus with us. 



Distrihidion : Appears to be widely di.stributed along the 

 northern F^uropean coa.sts of the .\tlantic and its arms. Rather 

 coninioii on the coasts of Norway, both on the west and north. 

 Occurs right up to the arctic coasts but appears chietly to be a 

 temperate species. 



C. concinnus \\ . Sm. 



The characteristic form of the living cell is illustrated by 

 OsTEXi'EM) (L. 116, p. 566). Its structure is always very fine, 

 much finer than usual in C. centralis. The central rosette has very 

 large areoles. 



As a distinguishing feature from C', centrulis the areoles in the 

 central rosette are several times larger than the others, those in 

 C. centrdlis being only twice as large, or sometimes but very little 

 larger. I am. however, not sure if this ditt'erence is always to be found. 

 Besides, the areoles in C', concinnus are „little marked" (cfr. Grax 

 L. 70, p. 168) with comparatively broad hyaline spaces between them, 

 and from the mai-ginal apiculi thei-e are similar rather broad, hyaline 

 stripes radiating inwards at some length on the valve. These 

 hyaline radii give the valve a highly characteristic appearance, which 

 C. centralis does not possess. 



Distrilnttion : On the whole, the same as that of C', centralis, but 

 seems to be still more extended. Seems to be rare on the arctic 

 coasts. On the west coast of Nonvay frequent in spring and 

 autumn, especially in the spring inflow. On the northern coasts of 

 Norway rare, at any rate in the months January — May. 



