REPORT OX THE BOTANY OF THE ISLANDS OF THE SOUTHERN OCEAN. 159 



Carex insularis, Carmich. 



Carex insularis, Carmich. in Trans. Linn. Soc. Lond., xii. p. 508; Boott, Carices, iii., p. Ill, t. 350. 

 Carex spic-is pluribns masculis femineis pedunculatis, Thouars, Esquisse Fl. Trist., p. 36. 



Tristan da Cunha. — In the plain — Carmichael. Inaccessible Island. Moseley. 

 Endemic in the group. 



Carex thouarsii, Carmich. 



Carex thouarsii, Carmich. in Trans. Linn. Soc. Lond., xii. p. 508; Boott, Carices, iv., p. 176, t. 595. 

 Carex spicis confertis sessilibus hermaphroditis, Thouars, Esquisse, Fl. Trist., p. 36. 



Tristan da Cunha. — In the plain — Carmichael; MacGillivray. Nightingale 

 Island. Moseley. 



Endemic in the group. 



Uncinia brevicaulis, Thouars, var. robustior, Hemsl. (Plate XLV.) 



Uncinia brevicaidis, Thouars, Esquisse Fl. Trist., p. 35, t. 6, var. robustior, Hemsl. 

 Uncinia brevicuhnis, Carmich. in Trans. Linn. Soc. Lond., xii. p. 508. 

 Uncinia gracilis /3, C. B. Clarke in Journ. Linn. Soc. Lond., xx. p. 400. 



Tristan da Cunha. Thouars; Carmichael. 



Also in St. Paul and Amsterdam Islands. 



Mr C. B. Clarke (Journ. Linn. Soc. Lond., xx. p. 400) combines Uncinia gracilis, 

 Thouars, with Uncinia breviculmis, Carmich. (the name Uncinia brevicaulis, Thouars, beino- 

 omitted) and Uncinia macloviana, Gaud., under the name gracilis, defining, however, three 

 varieties. We prefer keeping the South American form as a distinct species, because it 

 is easily recognised by its narrower utricles, which considerably overtop the bracts ; but 

 Thouars' two species can at the most be regarded as varieties of one. By some oversight 

 Mr Clarke has confused the two forms. He cites Carmichael's Uncinia brevicuhnis under 

 what he regards as typical Uncinia gracilis ; while his /3 gracilis is evidently a slip, and 

 should have been, according to Thouars' plate, quoted /3 brevicaulis. With regard to the 

 relative length of leaves and culms, Clarke says of his /3, " Culmi (in eadem plauta) foliis' 

 multo breviores aut midto longiores." The specimen in question bears the culms of two 

 seasons, one of the culms belonging the former season being longer than the present leaves ; 

 yet in all probability, judging from the other specimens of the species we have seen, the 

 leaves of the same season were longer than the culm. From some cause the growth of the 

 plant was not so vigorous the year the specimen was collected as it had been the year 

 before, hence the disparity in the length of the culms. Carmichael, loc. cit., doubtingly 

 unites Thouars' two species, and states that the culm greatly elongates after flowering. 

 This is so unlikely that we think he likewise may have been deceived by relative length 

 of previous years' culms. 



