I'lotistplankton. 



ijinal spines and plainer excentric secondai-y curves. Tlie figures 

 referred to above belong- to the latter form. There appears, how- 

 ever, to be so much variety, both in tlie nunibei' of tlie spines and 

 tlicir distance from each othci-, and in the numbei- of the fasciculi, 

 that it does not seem adviseablo to look upon the two series of forms 

 as bcin<;- specifically different. 



As the species does not always appear to be riirlitiy under- 

 stood, I iiavo added a few remarks on its structure. 



Around a central areole there are, as in t', cxcciitrinifi, 7 

 areoles forming, in conjunction with the central areole, an indistinct 

 central rosette. From here the areoles decrease in size (|uickly and 

 evenly right out to the margin. The secondary curves near the 

 margin are nearly straight, often neai'ly to "A of the radius reckoned 

 from the margin inwards, but further in, towards tlie centre, de- 

 cidedly concave outwards, as in C. excentricus. The valve is de- 

 cidedly convex on account of the high and abrupt marginal zone, 

 and is thus easily discerned from C. excentrimis, which is nearly 

 quite flat. Besides Thalassiosini decqAens always has the very long, 

 bent marginal spines. An odd, asymmetrical, spine is always 

 present. 



In side view the chains may very easily be taken for T. Nor- 

 dennholdi, whose structure, however, is altogether different. 



Only observed during the inflow of diatoms, at which time it 

 was abundant and frequently (especially in 1900) in large quantities. 



Distribution: Seems to be the same as that of Thcdassiosira 

 JS^rdeiiskioldi and T. yravida, and is often met with in their com- 

 pany. Yet, the secojidary maximum in August is wanting (on the 

 west coast of Norway and the Faeroes), and the species is, on the 

 whole, very rare except at the time of the Spring inflow. 



Co!!ieiiiowira Gean. 

 C. polycborda. (Gran) Gran. 



As I have mentioned in an earlier paper, (L. 92, p. 24), this 

 species may easily be confused with Coscinodiscus lineatm, as the 

 characteristic transverse processus are often difficult to discover, 

 and it seems possible that thy may be altogether wanting. The most 

 frequent form with 6 fasciculi corresponding to 6 transverse pro- 

 cessus •uill, thus, on the whole, have the same structure as Cose. 

 Vineatus, only much finer. The specimens -which I have tabulated 

 as Cose, linmtiis var. from a few places, are perhaps such forms 

 of Coseinosira poJyehorda where the transverse processus are wanting. 



Marginal apiculi, granules and various other processus on the 

 valves of diatoms seem to be rather inconstant, or at any rate very 

 varying, which probably is owing to theu- being more or less in- 

 completely silicated. 



Very rare, except at the time of the inflow of diatoms, when 

 it is abundant. 



Distribution: On the whole, the same as 2 halassiosira Nor- 

 denskioldi and T. grarula. It appears to occur all the year round 

 on the west coast of Norway, where it has, at any rate, been found 

 in most months; but it is common only during the inflow in the 

 Spring. 



Poro»«ira n. gen. 



Structure of the valve is in the only known species very fine, 

 in other respects as Podosira hormo'ldcs. Over tiie whole surface. 



thickest along the margin, scattered pores which probably are the 

 perforations of more or less plain short, hollow spines. Such ai-e 

 seen near tlie margin after destruction of the organic matter by 

 burning. 



Inside tiie margin at one spot on the surface, a large, strong, 

 odd (asymmetrical) spine. The connective zone is apparently for- 

 med of numerous rings, which, owin-j- to their delicacy, arc rather 

 indistinct. 



Forms chains, of two or a kw links, in which the latter are 

 joined jtogether by a short, and very thick, central mucilaginous 

 band, in wiiich one may with some difficulty discover fine threads. 

 The band seems almost to be structureless and is as good a,s in- 

 visible in water, but is easily seen on colouring with various dyes 

 e. g. methylene blue and gentian violet. 



Chromatophores, on the whole, the same as in Coscinodiscus, 

 polygonally roundish discs scattered along the valves and the con- 

 nective zone. 



It seems to me to be somewhat unnatural, like Gkan, to refer 

 the following species to the genus Lauderia. I think it would be 

 best to limit that genus to those species only which form stiff 

 chains of links which touch each other). On the other hand, the 

 difference between this species and Thcdassiosira is so great that 

 it should scarcely be considered as belonging to this genus either, 

 notwithstanding that the chains, generally speaking, are similarly 

 formed. There is too, according to my opinion, another important 

 difference in the structure of these genera, wliicii 1 hope to be 

 able to explain more fully on a later occasion. 



The genus Podosira forms stipitated chains (of 2 or a few in- 

 dividuals). Perhaps it will, however, prove not to be possible to 

 carry out a systematic classification based upon such principles, al- 

 though it would seem to be an important considei'ation, in a natural 

 system, whether a chain colony is swimming freely about (plantonic) 

 or is fettered. On the other hand it may perhaps be found that 

 several species of Coscinodiscus with scattered dots on the surface 

 wiW find a more natural place in my proposed genus Porosira. 



P. glacialis (Grun.) 

 (PI. VI, fio-. 7). 



Podosira hormoides var. glacialis Grux. L. 83, p. 56, pi. 5, 

 f. 32. P. glaeialis (Geun.) Cl. L. 27, p. 24. Lauderia glacicdis 

 (Geux.) Gran L. 68, p. 111. 



Out of tlie pores of the valve extend fine threads, which are 

 only seen with difficulty, in a very thick, short mucilaginous cy- 

 linder, which connects both valves. Probably similar threads ex- 

 tend without this cylindei". 



There are also near the margin, long fine, mucilaginous threads 

 which extend obliquely outwards and downwards, and probably 

 serve as a floating apparatus. Precisely similar threads are found 

 in Thalassiosira gravida, whose structure is remarkably like tliat 

 of the above species. 



Very scarce except during the inflow of diatoms, but then 

 very frequent and often in great quantities, especially in 1899. 



Distribution: Yet unsufficiently known, but probably, on the 

 whole, the same as Thalassiosira Nordenskioldi. Also found in the 

 Kara Sea. Numerous in the year 1900 during the inflow of dia- 

 toms in the Spring on the west coast of Norway. 



13 



