288 Transactions. — Botany. 



rhizomes and long stringy roots, it occurs in immense abund- 

 ance, often covering hundreds of acres to the exclusion of all 

 other vegetation. Mr. J. Stewart, C.E., informs me that the 

 workmen engaged in constructing the railway dreaded to 

 encounter it, as its thick matted roots not only made it 

 difficult to open out the drains, but were always a sure sign 

 of a very bad part in the swamp." 



Olearia semidentata varies in form according to the posi- 

 tion in which it is growing. When it forms a constituent of 

 the wettest portion of the Lepyrodia bog it is of a straggly 

 growth, its lower part leafless and concealed by the thick 

 mass of Lepyrodia, while only the ends of the branches, 

 which project into the light, bear leaves. When growing 

 where it has more room to spread it forms rounded bushes 

 sometimes 85 cm. in height and 1-24 m. in breadth. Such 

 plants, as I have pointed out elsewhere (12 s ), bear a rather 

 close resemblance to certain cultivated species of south Euro- 

 pean Cystus. The leaves are variable in size, thick and coria- 

 ceous, slightly cottony and shining on their upper surfaces ; 

 but the lower surface, which alone contains stomata, is thickly 

 covered with white tomentum. The ultimate branchlets are 

 so dense as to touch one another. Usually the flowers of 

 New Zealand plants are either white or yellow, but 0. 

 semidentata is a notable exception, the ray-florets being of 

 the most brilliant purple (see coloured plate, 23 2 , pi. ii.) ; 

 indeed, a group of these shrubs covered with heads of 

 blossom each 4 cm. in diameter is a most beautiful spec- 

 tacle. The roots, which are of considerable length, pro- 

 ject laterally, and not vertically, downwards. This manner 

 of growth is evidently strongly hereditary, since it ap- 

 pears in the seedling at quite an early stage. It would 

 be thought that a plant such as 0. semidentata, growing 

 under edaphic and climatic conditions of extreme con- 

 stancy, would vary little. On the contrary, so far as 

 the leaves are concerned, the amount of variation is very 

 great. I specially measured the leaves of a considerable 

 number of plants growing near one another in the vicinity of 

 Lake Eangatapu. The largest measured 6 - 5cm. by 1cm., 

 and the smallest 3 - 6cm. by 7 mm. The following measure- 

 ments give examples of the variability of length with regard to 

 breadth : 6'5 cm. by 1 cm. ; 5 - l cm. by 1 cm. ; 4 cm. by 1 cm. 

 The tomentum on the upper surface of the leaf varied, accord- 

 ing to my notes, from "none" to "very abundant." The 

 number of teeth on the margin are also very variable, both on 

 different leaves and on the opposite sides of the same leaf, 

 some leaves having no teeth at all and some as many as nine, 

 while others showed all the intermediate numbers. My notes 

 on this subject conclude thus : " Where I write, three plants 



