Paa den ene side er der mange arter, hvis kiselpanser 

 er saa tyndt, at det falder sammen og bliver ukjendeligt 

 ved indtorring, og selv jiaa de grovere arter kan cellevæg- 

 gen lide formforandringer, saa at karakteristiske forhold 

 undgaar opmerksomheden. Paa den anden side mister man 

 derved anledningen til at iagttage celleindholdet; dette 

 giver meget gode systematiske karakterer, som det er paa- 

 vist af Pfitzer (1871) og Schiltt (1888, 1895). 



Naar algcrne straks ved indsamlingen konserveres i 

 sterk alkohol, fikseres baade kromatoforer og cellekjerne; 

 de bør da helst larves f. eks. med safranin, som gjør meget 

 god tjeneste og er let at behandle, og undersoges i vand. 

 Med denne methode kan man naa længere end med nogen 

 anden hidtil anvendt; arterne har da et saa karakteristisk 

 udseende ved sin ydre form og sit celleindhold, at de kan 

 gjenkjendes selv ved temmelig svage forstørrelser. 



Knn naar man skal iindersøge hornenes eller livile- 

 sporernes struktur eller cellerne i sidestilling, kan det være 

 nødvendigt at l)enytte glødningsmethoden eller kemisk be- 

 handling. 



C/iætoceros-arternes cellevæg bestaar af to _sliaJIer 

 (valræ) og et iaand (tysk giirtelband) ; knn under au.xospore- 

 dannelsen har cellen to baand, et fæstet til hver af skal- 

 lerne. Baandet eller den forbindende membran har form 

 af en kortere eller længere aaben cylinder. Det er ved 

 sømmene (suturaé) forbundet med skallenes cylindriske 

 del, shalkappen. Ved sømmene er celle væggen oftest noget 

 indsnevret, undertiden temmelig sterkt (f. eks. hos Cliæto- 

 ceros constridum n. sp.). Skallernes Inmd har form af en 

 mere eller mindre langstrakt ellipse, som kan være hvælvet 

 eller konkav. Nær enderne af ellipsens længste axe er 

 hornene fæstede. Ellipsens længste axe kalder jeg med 

 Scliiitt (1895) sagittalaxen (s), den korte axe transversal- 

 axen (f). Den tredie axe, som gaar fra midten af den ene 

 skal til midten af den anden, kaldes loitgHudinalaxen (?)•*) 

 Det jjlan, som kan lægges gjennem denne og sagittalaxen, 

 kaldes sagittalplanet. 



Cellen siges at ligge i frontstilling**), naar baandets 

 bredeste side er vendt mod iagttageren (breite Giirtelband- 

 ansicht). Naar skallens flade sees, kaldes det sidestilling 

 (Schalenansicht), og naar baandets smaleste side sees, sa- 

 gittidst Uling (schmalle Giirtelbandansicht). 



Hornene udgaar et fra hvert af cellens 4 hjørner 

 (cellen tænkes i frontstilling); de binder cellerne sammen i 



*) Otto Muller (189.')) har for diatomaceernes axer indfprt anJre 

 navne, s = apikalaxen, t = traiisai:iikalasen, I = pervalvaraxen. 

 Da Cleve har benyttet Sehiitfa betegnelser, har jeg ogsaa troet 

 at burde følge dera. 

 **) Schiltt afi-iger her fra de ældre (eugelske) forfattere; jeg har 

 ment at bm-de folge disse i nomenklaturen, selv om Schiitts be- 

 tegnelser i og for sig synes mig mere hensigtsmæssige. 



There are, on thi' one hand. niany species whose 

 siliceous envelope is so thin that it collapses and becomes 

 iinrecoguizable after drying. aud even in the coarser species, 

 the cell wall may undergo changes of form, so that the 

 characteristic conditions escape notice. On the other hand, 

 the ojiportunity is thereby lost of observing the contents 

 of the cell, which exhibit very good systematic characters, 

 as demonstrated hy Pfitzer (1871) and Sciiiitt (1888, 

 1895). 



Whcn the algæ, immediately after collection, are pre- 

 served in alcohol, both tiie chromatophores and the cell- 

 nuelous are fixed, and should thcn be stained (e. g. with 

 satiranin, which answers the purpose very well, and is easy 

 to manipulate), and examined in water. By this method 

 we are enabled to get farther than by any other hitherto 

 employed; the external form and cellular contents of the 

 species are then so characteristic in appearance, that tbey 

 are recognisable even under a tolerahly weak magnifying 

 power. 



Only when the structure of the setæ or the spores, 

 or a side view of the cells is to be examined, may it be 

 necessary to expose to the action of fire or to boil with 

 acids. 



The cell wall of the Cliætoceros siiecies consists of 

 two valres (^valvæ) and a lioop (German, Giirtelband). 

 During the formation of the auxospores the cell has two 

 hoops, one connected to each of the valves. The hoop, 

 or connecting membrane has the shape of shorter or 

 longer of)en cylinder. At the sutures (sidiirac), it is 

 connected with the cylindrical portion of the valves, the 

 mantle. At the sutures, the cell wall is often narrowed. 

 sometimes very considerably (e. g. in Cliætoceros constrictum 

 n. sp.). The bottom of the valves has the form of a more 

 or less elongated ellipse, which may be convex or concave. 

 The setæ are attached near the extremities of the long 

 axis of the ellipse. In common with IScJiUtt (1895), I call 

 the long axis of the ellipse the sagittal axis (.<?), and the 

 short axis, the fransrerse axis (t) ; the third axis, which 

 goes from the middle of one valve to the middle of the- 

 other, is called the longitudinal axis*). The plane which 

 may be made to pass through this and the sagittal axis, 

 is called the sagittal plane. 



That view in which the hroadest side of the hoop is 

 turned towards the observer is termed tiie froid rieiv**) 

 of the cell (breite Giirtelbandansicht). When the surface 

 of the valve is seen, it is called the side vieiv (Schalenan- 

 sicht); and when the narrowest side of the hoop is seen, 

 the sagittal view (schmalle Giirtelbandansicht). 



The setæ issue one from each of the four angles of 

 the cell (the front view of the cell is the one here sup- 



*) Otto Muller (1895) has given other uames to the axes of the 

 Diatomaceæ: s = apical axis, f = transapical axis, i = pervalvar 

 axis. As Cleve employs SchUffs terminology, I have considered 

 it right to do the same. 

 **) .Schutt here differs from the earlier (English) writers, I have 

 considered it right to follow the latter in their nomenclature, 

 even thoiigh Schiitfs designations in themselves may be more 

 convenient. 



