1136 APPENDIX. 
Page 
109 
112 
120 
121 
128 
128 
131 
total absence of resin-canals, which are present in all the 
other genera of dnacardiacee, but considers that this pecu- 
larity is not accompanied by correlated characters of sufficient 
importance to justify the exclusion of the genus from the order. 
The discovery of Corynocarpus in western Polynesia is of con- 
siderable interest in connection with the often-quoted tradition 
that the New Zealand species was introduced by the Maoris 
when they first colonised the country. 
XXII. LEGUMINOS£. 
Carmichaelia Muelleriana, [egel in Garienf. (1887) 611. 
—I have been unable to satisfactorily identify this plant, which 
is probably the same as one of the species described in this 
work. 
C. Williamsii.—Barren specimens apparently referable to 
this species have been collected by Dr. Cockayne on the Poor 
Knights Islands, but the identification cannot be considered 
certain until flowering and fruiting specimens have been 
obtained. 
Clianthus puniceus.—A variety with white flowers is now 
commonly cultivated in gardens. 
Swainsona nove-zealandize.— Mount Torlesse Range, 
Canterbury, alt. 3200-3600 ft., Dr. Cockayne. 
XXIII. ROSACEZ:. 
Geum aucklandicum, Greene, Pittonia, iv. 225.—This 
name must be substituted for G. sericewm, T. Kirk, over which 
it has several years’ priority. 
G. uniflorum.— Dr. Cockayne remarks that this is an 
extremely common plant of subalpine and alpine meadows and 
moist shady rocky places in Westland. 
Aceena sanguisorbe var. antarctica, Cockayne in Trans. 
N.Z. Inst. xxxvi. (1904) 319.— Dr. Cockayne proposes this 
name for the common form of the species in the Auckland 
and Campbell Islands, remarking that it can be distinguished 
from var. pilosa by the pale-green (not whitish-green) leaves 
which are glabrous on the upper surface, much shorter scapes 
which are hardly raised above the foliage, bright-green calyx- 
lobes, and by the shorter stamens. I have seen no specimens. 
