78 ON EPILOBIUM DURI2BI. 



Hab. Chili. Carrizalbayo, F. Philippi £ Borchers, 1885 ! 

 Stem |-2 ft. ; leaves 1-1^ in. ; sepals J in. ; petals f in. 



40. C. Molina Gay, Fl. Chile, i. p. 326. 

 Hab. Chili. Northern Provinces ! 



41. C. ranunculifolia Phil, in Herb. Mus. Brit, ined.— Caule 

 erecto ramoso folioso, foliis petiolatis luteo-viridis palmatilobatis 

 vel serratis basi cordatis minute pubescentibus, floribus paniculatis 

 parvis, sepalis triangularibus acutis, petalis caeruleis calyce duplo 

 longioribus, carpellis ignotis. 



Hab. Chili. Atacama, F. Philippi 4 Borchers, 1885 ! 

 Stem 1-1 J ft. ; leaves 1 in. long ; sepals J in. ; petals ^ in. 



Species exclasa. 

 C< coccinea Pursh. = Mai vast rum coccineum A. Grav. 



Non satis nota. 



iliversifolia 

 Haenkeana 



C. insularis F. Ph. in Anales Univ. Chil. 1875, p. 187. 

 C. patcujonica Phil, in Linnsea, xxxiii. p. 28. 



C. pauciflora Poepp. in Hb. Mus. Brit. ; material insufficient for 

 determination. 



(To be continued.) 



ON EPILOBIUM DTJTilMI J. Gay 



By C. B. Clakke, F.B.S. 



Mr 



Journ. Bat. 1891, pp. 225-228, t. 307. Mr. Marshall identifies my 

 English specimens as E. montanum Linn., forma minor aprica 

 Hausskn., to which I have no objection; but he proceeds to say 

 that E. Duriai J. Gay is an altogether different plant ; and on p. 297 

 he gives the five points of difference on which he relies, as together 

 amounting to a specific difference. 



^ I should not attempt to encumber your pages further if the 

 point at issue was merely whether a plant considered by one 

 botanist a species was in the eyes of another a var. It is quite 

 possible that Haussknecht or Marshall could sort success- 

 fully specimens of E. Dnriai of the South of France from 

 specimens of E. mmtomm, forma aprica, of the South of England. 

 So in many cases Mr. Baker can separate his " north-country 



form" from my Hampshire examples of what is universally called 

 the same species. 



But Mr. Marshall appears to me not to have realised the 

 strength of the case I originally put together. I show that the 

 characters relied on by Haussknecht and his ordering of the forms 

 differ from those adopted by Nyman and other high authorities ; I 

 show that the differences in the seeds, on which Haussknecht 



