BY ALEX G. HAMILTON. 767 



(7) The inner cuticle (5/(.). Through this penetrate uniceUuLar 

 hairs (iig. 10) which are epidermal outgrowths, and are thick- 

 walled and destitute of contents. They are rarely septate as 

 shown in the figure, but usually resemble those of Coprosma lucida 

 (fig. 11). This cuticle, as above remarked, has no storaata. 



The same layers, omitting the palisade-parenchyma, are met 

 with in the floor of the cavity, but in i^eversed order, and in the 

 rim is a vascular bundle composed of five or six vessels. 



On examining leaves of various stages of growth, I find that in 

 leaves 5-9 mm. in length, the domatia appear as slight hollows. 

 In leaves 1-9 cm. long I find the hollow deeper, and a little tuft of 

 hairs in the angle. These are of two kinds : the ordinary pointed 

 hair (fig. 10) and short thick ones composed of four almost 

 globular cells. In a leaf of 4 cm. long a thickening is apparent 

 along the sides of the veins, making a triangular pocket as in 

 Group ii., and the hairs project from this. At 5 cm. long the 

 thickening begins to extend across the mouth from the sides, so 

 that there is a hollow surrounded by a ridge. Up to this stage 

 the whole of the under side of the leaf is a purplish-brown in 

 colour, bvit the ridge is a very bright green. The ridge had 

 grown higher all round in leaves 5-5 cm. long, and a few hairs 

 had grown on the front part of the ridge, their points directed 

 towards the centre of the hollow. In leaves 6 cm. long the greater 

 height makes the cavity appear much deeper. At 8 cm. the ridge 

 has reached its full height, and there are a few hairs on the out- 

 side of the ridge — simple and pointed. The domatia are com- 

 pletely formed when the leaf is 11 cm. long, and no further 

 alteration takes place except that in leaves a year old there are 

 fewer hairs in the interior of the domatium. '^'he leaves reach a 

 length, when full grown, of 16 cm. and upwards. In examining a 

 large series of young leaves, I found no Acari present until the 

 domatium was fully formed. This fact has an important bearing 

 on Dr. Lundstrom's theory of the meaning of the structures. 



Coprosma lucida, Forst.- — This plant also belongs to Group i. 

 The domatia are very large and highly developed. They occur 

 in the axils of the secondary veins and midrib, in pairs, or 



