46 F. E. FRITSCH. 



BACILLAEIEJE. 



COSCINODISCACE^E. 



Genus MELOSIRA Ag. 



67. MELOSIRA sp. 

 (PI. III., figs. 148, 149, 149a.) 



Diam. cell = 7-9 p. ; long. cell. = 12-13/4. 



Hull. On Phormidium on ice, four feet above frozen watercourse through 

 "Penknife ice," McMurclo Bay, September 13th, 1902. 



I have not been able to come to any satisfactory conclusions about the 

 determination of this form. It occurred rarely (but then in considerable masses) on 

 the surface of Phormidium at the above locality, and was always found dead. The 

 valves were, in most cases, hemispherical, perfectly smooth (i.e. without markings of 

 any kind), and there was generally a prominent constriction at the point of junction 

 of the two valves (cf. the figures). Occasionally the line of junction of the valves 

 showed a punctate structure like a number of minute interlocking teeth (fig. 149^). 

 Frequently the length of the valve was nearly equal to its breadth (fig. 14 ( J), but 

 sometimes the cells were appreciably longer (fig. 148) than wide. 



The frustules were generally isolated, as in the section Pudi>sira of the genus ; 

 but it does not appear to be related to any of the described species of that section. 

 It is most like P. nn>nt<i<ini'i Kiitz., a much larger form, which has markings on its 

 valves. It may well be, however, that the frustules had fallen apart in the dead 

 material, and that our form belongs to the section La/sigonium, 



The characters of the frustules agree fairly well with the descriptions of 

 M. subftexills Kiitz., given by Rabeuhorst (Fl. Europ. Alg., i. (1864), p. 39) and 

 Migula (Kryptogamenflora v. Deutschlaud, Deutsch-Osterreich, etc., ii., 1 (1907), 

 p. 153), but Kiitzing's figure (Bacill. (1844), Tab. 2, fig. xiii.) shows the valves as far 

 less convex than in the Antarctic form ; there is also practically no constriction and 

 no indication of teeth at the line of junction of the valves (cf. op. cit., Tab. 2, fig. vi. of 

 M. ifdlica). Species like J/. lineata (Dillw.) Ag., and M. salina Kiitz., should also be 

 compared. It is possible that the Antarctic form is a new species, but the material 

 was too inadequate to enable one to arrive at a definite conclusion on this point. 



