332 Transactions. — Botany. 



veins. But this state is only to be seen in sheltered situa- 

 tions on the hills where nutritive matter is abundant and the 

 plant has a sufficiency of saline moisture without being 

 saturated by heavy spray. 



In forma carina the stems are more robust, trailing or 

 prostrate ; the leaves are large, fleshy, and shining ; while the 

 flowers are but sparingly produced, are of small size, and 

 somewhat hidden amongst the luxuriant leaves. This form is 

 found only amongst rocks often washed by waves, and con- 

 stantly subjected to heavy spray. If growing a few feet 

 above the actual reach of the waves, but still subject to heavy 

 spray, the prostrate branches are much shorter, the glossy 

 leaves are more closely set, while the flowers are larger and 

 more brightly coloured. 



In forma concinna the stems are short and crowded, never 

 exceeding oin. in length ; the leaves are small and close-set, 

 with solitary minute axillary purple or white flowers in the 

 uppermost axils. Charming compact little specimens less 

 than lin. in height may often be found bearing scores of 

 minute flowers, which are but rarely brightly coloured. This 

 form is found chiefly on the surface of huge masses of 

 Trichocolca tomentella and other Hepaticae, often in vast 

 abundance, in situations sheltered by a close scrubby growth 

 of various shrubs and trees, but still exposed to the finely- 

 divided sea-spray which constantly impregnates the atmo- 

 sphere of the Auckland Islands. Hence its annual duration, 

 diminutive size, crowded branches, close-set leaves, and 

 minute flowers are chiefly due to a lack of mineral consti- 

 tuents in its nutriment, and possibly, at certain times, to a 

 lack of moisture also. 



All these forms are connected by a series of insensible 

 gradations, so that it is impossible to say at any given point 

 in the series, " Here is a line of division." The change is so 

 gradual that, although it can be easily made out, it is impos- 

 sible to say where one form begins and another ends. 



All the New Zealand species agree in having a five-cleft 

 calyx, a monopetalous corolla more or less deeply divided 

 into five segments which are never barbed, five epipetalous 

 stamens with small versatile anthers, and one-celled two- 

 valved many-seeded capsules. 



Their local distribution is somewhat remarkable, as they 

 are not represented in the extreme northern part of the colony, 

 their limit in that direction being attained on Hikurangi, a 

 little to the north of the 38th parallel, from whence they ex- 

 tend southward to the Chatham Islands, Stewart Island, An- 

 tipodes Island, the Auckland and Campbell Islands, but do not 

 reach Macquarie Island, thus illustrating the law previously 

 observed in the Northern Hemisphere that, although plenti- 



