300 



Part II I. — Fifteenth Annual Report 



Gosdnodisctts oculus iridis, Ehb. A single specimen only was found 

 in the mixed sample. 



Dactylimolen antareticus, Castr. (Challenger D., p. 75, PI. IX. fig. 7), 

 occurs in several samples, but nowhere abundant. Hitherto observed in 

 the Antarctic Ocean only. 



Dactyliosolen mediterraneus, H. P. var. tenuis, CI. (PI. fig. 14 ). 

 Frustulea slightly silicious, in diameter 001 to 0*025 mm. Annuli, 2 to 4 

 in 0*01 mm. Valve flat, with delicate puncta arranged in rows, which 

 radiate from the centre, where are some scattered puncta. Connective 

 zone, with fine, elongate puncta arranged in lines parallel to the 

 longitudinal axis, about 24 in O'Ol mm. 



This form agrees with D. medit, H. P. (Monogr. d. Ehizosol, PL I. figs. 

 8, 9), in most respects, but not in the strength of the membrane, which is 

 described as being strongly silicious. 



Hemidiscus cuneiformis, Wallich (Janisch. Gazelle Exp., PI. I. 

 fig. 7). Some few valves in the mixed sample. 



Navicula fusiformis, Grun. var. A very small, exceedingly delicate, 

 form, in length 0*035 mm. only, was found in the mixed sample. 



Nitzsehia closterium, W. Sm. A delicate variety, with straight beaks, 

 was rarely found in Nos. 5, 8, 12, 16, 17. 



Nitzsehia lineola, CI., n. sp. (PI. fig. 10). Valve narrow, acute. 

 L., 0*1 to 0*11 mm.; B., 0*002 mm. Keel eccentric, its puncta 14 in 

 0*01 mm. Striae, 24 in 0*01 mm. This exceedingly delicate form was 

 observed rarely in No. 14. 



Nitzsehia {Tryblionella) migrans, CI., n. sp. (PI. fig. 9). Valve 

 flat, with parallel margins and suhcuneate somewhat obtuse ends. L., 

 0*035 mm. ; B., 0*007 mm. Keel-puncta indistinct. Striae coarse, 11 in 

 0*01 mm. 



This species was found in isolated specimens, rarely in Nos. 12 and 13. 

 It is related to N. seriata, CI. 



Nitzsehia seriata, CI. (Vega Exp. D., PI. XXXVIII. fig. 75). This 

 species, which occurs in the Arctic Sea, and occasionally in the North Sea 

 the Irish Channel, was met with in some few samples only. It is the 

 only arctic form observed in the ' Research ' samples. 



This species may easily be mistaken for another form, which I name 

 N. fraudulenta, n. sp. The latter (PL fig. 11) occurs in chains as 

 N. seriata, but is much more delicate, and has finer striae (23 in 0*01 

 mm.), which are seen only with difficulty, whilst N. seriata is coarsely 

 striate. 



N. fraudulenta was found last summer in the harbour of Plymouth, 

 and at the end of October in the middle of the North Sea. It is a 

 southern, and N. seriata a northern, form, perhaps of the same species. 



Rhizosolenia alata, Btw. (M.J. VI., PL V. fig. 8). This species was 

 exceedingly common in several samples, especially from the western 

 stations. The form found there was the typical one, but variable in size, 

 so that several specimens must be considered as belonging to R. alata 

 var. gracillima, CI. The latter form, but not the typical, occurs in 

 summer-time in large masses in the Kattegatt and Skagerak. This 

 variety was also found as the principal constituent of the plankton near 

 Beeren Eiland in September this year. At both places I have sometimes 

 met with specimens, of which one-half belongs to R. alata and the other 

 to R. gracillima (see the fig. in Van Heurck Synopsis, PL LXXIX. 

 fig. 8). I am very much inclined to consider R. alata as an Atlantic 

 form, which, in water of less salinity, becomes changed into R. gracillima. 



Rhizosolenia semispina, Hensen (PL fig. 13). Under the name R. 

 setigera, two well distinct species have been confounded, viz., the original 



