ion 



rgen<eii. 



ved in formaline, wiien tliey liave not been ignited on cover-glass; 

 but tliey may be wanting and sometimes be alsp very distinct in one 

 and the same chain. It appears to me, however, that they become 

 smaller and may even disappear altogether during the process of 

 igniting. They are, howevei', also often wanting in specimens pre- 

 served in formaline. 



It has not hitherto been possible to see any definite structure. 

 Cleve, who considers this species to be identical with Osteup's 

 X. fcidentnoncdis. has also mentioned that he has not been able 

 to dissolve the structure. I have examined numerous specimens, 

 most of them in side view, and have noticed a narrow linear stau- 

 roslike marking, which, however, was very indistinct, but I have 

 not been able to discover anything further. 



Rather freijuent during the diatom inflow in the spring, some- 

 times numerous; otherwise wanting. 



Distribution: Arctic coast form, known from Greenland and 

 the Barents Sea (Heimdal 1900, cfr. Gran L. 70, p. 182). Accord- 

 ing to Gran (1. c.) N. Vanhuffeni was very frequent in the Siger 

 Fiord in Yesteraalen "A 1901. Not observed on the west coast of 

 Norway. According to Cleve (L. 40, p. 333) also found in the 

 r.altic. 



N. pelagica Cl. 

 (PI. Xn, fig. 23). 



Cl. L. 26, p. 11, pi. 1, f. 9. 



Was not so very I'ai'e in the plankton 1899. Only found in 

 1900 in one of the samples: The Salten Fiord, V4, 0—330 m. 



I have never succeeded in finding this species in material pre- 

 served in alcohol oi' formaline. I only know it from slides, where 

 it has been ignited on cover-glass, and have never seen it in valvar j 

 view. It answers fully to Cleve's description and illustration, so 

 that there can hardly be room for any doubt as to its being cor- 

 rectly determined. 



The peculiar „hairs" are, according to ray opinion, fragments 

 of the complex connecting zone and probably appear only after 

 being ignited on cover-glass (or treated with acids, which, however, 

 seems to destroy the cells). Is certainly very closely related to 

 N. Vanhoffeni. 



Distrihutiun : Arctic species, probably, like the foregoing, a 

 coast form. Hitherto only known from Baffin's Bay, Davis Strait 

 and the J3arents Sea (S/S Heimdal ''A 1900, r). 



Ktanroiieiis septentrionalia Gkun. 

 (1=1. vir, fij.-. 24). 



Geunow L. 83, p. 53, pi. 1, f. 48. Navicula septentrionalis 

 Oestk. I^. 139. p. 317, non Cl. L. 40; nee Cl. L. 48, p. 3, De 

 TuNi 1^. 50, p. 126, A. Schmidt L. 128, pi. 6, f. 37. Lihellus? 

 seytentnonalis Oestu. L. 138, p. 439, pi. S, t'. 97. 



This species does not occur in our noi'tliorn plankton. As be- 

 fore mentioned, the species which Cleve several times (e. g. L. 

 40) mentions as Navicula neptentrionalis Oeste., is identical with N. 

 Vanhoffeni Gean, at any rate, in part. Judging from his remark 

 on the contents of the cell (L. 65, p. 9) it would seem that Gean 

 also considers a form whicii is closely related to A^. Vanhoffeni to 

 be N. se}Aentrionalis Oeste., as he describes the contents of the 

 cell in both cases as being the same. Stauroneis septentrionalis 

 Gkun. has, however, quite another inner construction (v. below). 



As mentioned under the foreg-oing species, I earlier also conside- 



red a form to bo Xaricula septentrwnal'is, which I afterwards found 

 impossible to definitely distinguish from N. Vanhdffmi. Prepara- 

 tions ignited on cover-glass (not treated with acids) always answer 

 well to the description of N. sejjtentrionalis, as the central nodule 

 then becomes very distinct, and the spaces between the cells — as 

 before mentioned — generally disappear. 



A careful examination of the ends of the valve — cfr. pi. 

 VII, fig. 22 and 24 — will show that there is, however, here a 

 question of two very ditferent species, for Oesteup's illustration never, 

 in this respect, corresponds to the appearance of N. Vanhoffeni. 

 as I have had ample opportunity of seing in numerous preparations 

 of the latter. 



It is easily explained that Oesteup's species might be confused 

 with i\'. Vanhoffeni. as he — although very much in doubt, as he 

 has himself observed — refers the species to the genus Lihellus. 

 Soon after he changes the name of the genus, probably because 

 Cleve in the meanwhile (L. 26) has referred it to Xavicida. 

 Oesteup neither illustrates nor mentions the connecting zone as 

 being complex. As, however, the almost simultaneously discovered 

 N. Vanhoffeni has a distinct complex connective zone, and was 

 also found to occur as a pelagic species in long chains, it was very 

 easy to confound these two species. 



For the first time I was aware that a species exists, which 

 answers very precisely to Oesteup's drawing, by the pi'eviously 

 mentioned plankton sample from the Barents Sea (S/S Heimdal, 

 "/o 1900). Gean mentions this sample in his last work (L. 70, 

 p. 147 j and enters both K. Vanhoffeni and A. septentrionalis 

 Oeste. from this place. There was, however, in this sample, a 

 larger species with the same characteristic chromatophores as in 

 N. Vanhoffeni and also with a distinct stauros, which was easily 

 seen in watei'. As far as I can see, this must be a new species, 

 which I have described below. 



That the species represented on pi. N'll, f. 24, is the same as 

 Stauroneis sepAentrionalis Geun., I see no reason to doubt, the 

 more so as they were both found in the same waters. (Geunow's 

 species was found on the ice on the west of Novaja Semlja). The 

 species seems to have a partiality for the neighbourhood of ice. 



As it has not previously been found as a pelagic species in 

 chains, I adjoin the following short description. 



N'alve linear, more or less distinctly cuneate towards the ends, 

 which are either broad and rounded or almost square. Length 

 22—27 II., width 4—51/2 \^. 



Thei'e is a distinct central stauros, \\hich does not reach right 

 out to the sides of the valve, bounded by two lines, which are 

 parallel nearly out to the edge, where tliey ai'e clearly divergent. 

 Between them near the margin are 1 — 2 short, coarse striæ, simil- 

 arly radiating. The rest of the valve is transversely striated; the 

 striæ, however, are only seen with difficulty, with the exception of 

 those which are more widely separated in the middle. Forms long, 

 firm chains where the cells lie very close to each other, also at 

 the corners. On being ignited on cover-glass the chains break and 

 the ends of the cells become slightly separated from each otiicr. 

 (Cfr. pi. VII, f. 24.) 



Contents of the cell: As far as can be seen from the pre- 

 served specimens, from the central ])rotoplasm mass in which the 

 imcleus lies extend upper and lower arms (probably 4 in all. 2 

 upper on either side and two lower). Whether these unite to form 

 the usual chromatophores A\liicli ai'c found on the connecting zone 

 in the genus Navicula, I have not been a1:)le to decide. But the 



