OK DIATOMACEOUS GATHERINGS FROM KERGTJELEN'S LAN IX 55 



XXVIII. On the Diatomaceous Gatherings made at Kerguelen's 



elet, M.A., H.M.S. l Challenger.' By 

 , M,A. Communicated by Dr. Hooker, 



Mo 



Pres.R.S., V.RL.S. 



(Plate T.) 

 [Head April 15, 1875.] 



The gatherings were seven in number — four from fresh water, 

 two marine, and one from a freshwater lake within the reach of 



tidal influence. 



No. 1. Pond at sea-level, Tristan d'Acunha. This gathering, 

 as might have been anticipated, contained both freshwater and 

 marine forms: of the latter, Melosira moniliformis, Kiitz., abounded ; 

 and of the former there were numerous specimens of Fragilaria 

 vzrescens, Ralfs, F. capucina, Desm., Gomphonema tenelhim, Kiitz., 

 Gomphonema intricatum, Kiitz. 



No. 2. From a stream on the cliffs, Tristan d'Acunha. 5n 

 this were found Navicula viridis, Nitzsh, N carussias, Ehr., Fra- 

 gilaria capucina, Desm., F. virescens, Kalfs, F. cequalis, Heiberg, 

 F. crotonensu, Kitton. 



No. 3. From a stream, Tristan d'Acunha. This gathering con- 

 tained Achnanthidivm coarctatum, Breb., Fragilaria virescens, 

 Ralfs, Gomplionema calcareum, Clene, Navicula viridis, Nitzsh, 

 and N carassius, Ehr. 



No. 4. Small pond-streams, Marion Island. The forms found 

 in this gathering were Navicula carassius, Ehr., Diatoma grande, 

 ¥m. Sm., both in great abundance, and also an undescribed spe- 

 cies of Gomphonema^ which I propose to distinguish as 



Gomphonema bicapitatum. PI. L fig. 1. 



Frustules '0018 in length. On front view cuneate, slightly arched on 

 the margin. On the side view bicapitate. The attached end much 

 smaller than the opposite ; striae fine, costate, radiate, reaching the 

 median line. 



No. 5. Small pond-streams, Marion Island. This gathering, 

 though from a place similar to that of the preceding, is quite di- 

 stinct, containing in abundance Navicula carassius, Ehr., Fragilaria 

 capucina, Desm., Diatoma grande, Wm. Sm., and an undescribed 

 species of Diatoma which I would name as 



Diatoma rhombicum. PI. I. fig. 2. 



Frustules small, about '0006 in length. On front view quadrangular, 



LTNN. JOURX. BOTANY, VOL. XV. * 



