FRESHWATER ALG^E. 49 



TABELLARIACEiE. 



Genus Denticula Kiitz. 



76. Dentioula tenuis, 

 Var. ANTARCTICA var. nov. 



(PI. TIL, figs. 156, 157.) 



Dentkula tenuis Kiitz., Bacill. (1844), p, 43, t. 17, fig. viii. ; Smith, Brit. Diatom., ii. (18.56), p. 20, 

 PI. XXXIV., fig. 293. 



D. parva valvis late ellipticis apicibus rotundatis ; costis transversis validis, 

 6-7 in 10 ju,, in media parte rectis, sed polos versus gradatim plus minusve curvatis ; 

 inter binas costas sunt duse series punctarum subtilium, circa 6-7 in quaque serie in 

 media parte valvarum ; polos versus punctse in quaque serie fiunt minus crebrse. 



Long. valv. = 18-23 ju; lat. valv. = 7-8 /a. 



Ilab. — Pond some distance behind hut, Cape Adare, January 9th, 1902 ; on ice, 

 four feet above frozen watercourse through " Penknife ice," McMurdo Bay, September 

 13th, 1902. 



This was a relatively rare form, of which not many specimens were seen.* The 

 new variety resembles var. frig Ida Grun. {= D. frigida Kiitz.) in some respects, 

 but in the latter the number of eostse is not as great as in var. antarctica, the costse 

 are not curved near the poles of the valves, and there are more than two rows of 

 dots between each pair of ribs. There is also resemblance to var. inflata Grun. 

 {= D. inflata W. Smith), in which Smith {op. cit., PI. XXXIV., fig. 294) figures 

 many of the costse as curved (c/., however. Van Heurck, Synopsis Diat. Belg. (1880-85), 

 PL XLIX., figs. 32, 38), but the curvature is diff'erent, as the convex surface faces 

 towards the poles, whereas in var. antarctica it is the concave surface. 



In some of the valves the costse.near the two ends were curved more towards one 

 side than the other (fig. 156), leading to a certain degree of asymmetry. 



* The form here described bears an appreciable resemblance to the figure oiFragilaria antarctica Castracane, 

 given by Castracane ('GhalL' Bep., Diat. (1886), PI. XXV., fig. 12); this figure shows the same characteristic 

 paired rows of dots between each two ribs, and the general shape of the valve is also the same. The ribs are, 

 however, much more delicate and are scarcely curved at the ends of the valves. Van Heurck's figure (op. cit., 

 1909, PL III., fig. 48) shows nothing of the punctulations. Some doubt may be felt as to whether the valve figured 

 by Castracane really belongs to the genus Fragilaria. 



N 2 



