202 



SUBANTARCTIC ISLANDS OF NEW ZEALAND. 



[Ecoloi/lcill Bntillljl. 



Besides cases of variation according to environment, tliere are others wliere 

 distinct variations occur which cannot be referred to such, new characters appearing 

 in certain individuals which may be hereditary. The case of Celmdsia vernicosa 

 might belong either to this class or to the foregoing. It exhibits two distinct forms 

 growing in close juxtaposition — the one, roughly speaking, has an open rosette 

 with long flat leaves ; and the other a close rosette with shorter thicker leaves, 

 having a knobby apex and thickened incurved margins. Certainly, shade-conditions 

 niodify the rosettes. My notes are somewhat contradictory : e.g. — Adams Island : 

 " There are certainly two forms, the one with narrow leaves {i.e., T suppose, the 



I'lii. 7. — i)i;oiii'iN(i Haeit of the Tussock-Gkass I'oit Inllt 



knobby-apex form) being the plant par excellence of the highest points." Hills at 

 head of Carnley Harbour, on Auckland Island : " This plant varies so much accord- 

 ing to station as to look like different species, and, as the smaller and wind-swept 

 are the more frequent, it is the large specimens which appear abnormal." 



There are a number of interesting colour-variations. 8ome of the species are 

 well known to be variable in this regard ; in others the colour is considered one 

 of the specific marks, and the variations have not been recorded previously. 



Gentiana cerina, well known as variable in colour, was noted with the following 

 variations : (1) White ground-colour, stained rosy-purple at base, and with lines 

 of this colour extending upwards ; (2) white ground-colour, unstained at base, and 



