288 Willis . — The Floras of the Outlying Islands of 
from Stewart, Rhamnaceae, Anacardiaceae, Leguminosae, Solanaceae, Myo- 
poraceae, Piperaceae,and Palmaceae,are all northern except perhaps Legumi- 
nosae, which is the largest New Zealand family not yet recorded from 
Stewart. 
(30) In the same way. we shall expect the smaller genera of the 
Stewart flora to be those that are chiefly missing, and on examination we 
find that of those represented in Stewart by one species 57 are not found in 
the Chathams, and 31 have been found, while of those represented by two 
or three species 19 have not been found, and 26 are present, in the 
Chathams. 
(31) One may, by the aid of age and area, go a good way towards 
a prediction of the whole flora of the Chathams (cf. that of the flora of 
Stewart in 11 , p. 27). If one first predict that they will contain the 
Kermadec species that reach Dunedin, this gives 38 species of which 26 
reach the Chathams ; then adding that they will contain the Auckland 
species that reach Auckland city gives about as many more. If now one 
add to this that the bulk of the rest of the flora will likely be made up from 
the species which range the whole length of New Zealand (for those that 
ranged less distance would as a rule be too young to reach the Chathams at 
all, unless they ranged beyond New Zealand to the Kermadecs or Auck- 
lands), one obtains almost all the remaining flora, except the local endemics, 
which we have already seen may be to some extent predicted as likely to 
belong to the oldest families in the Chathams, that is to say, to the families 
represented by the most species in New Zealand (see above, p. 270). Of 
course this prediction that the Chatham ' flora will be selected almost 
entirely from Classes 1 and 2 (6, p.449) of the New Zealand flora brings in 
also a great many other species belonging to those classes, but which do not 
occur in the Chathams. We have as yet no means of deciding which are 
the oldest species in a given class, and it will only be the oldest, in general, 
which will reach the Chathams. The fact remains, however, that by this 
prediction we cover all of the Chatham flora but 14 species and the local 
endemics, and we have seen that we can more or less closely predict the 
families and genera to which these belong. These unpredicted species are : 
Pomaderris apetala : a special case, see 8, p. 332. 
Corynocarpas laevigatas : see above, p. 279. 
Tillaea moschata\ a coast species. 
Epilohium insulate : lowland swamps. 
Coprosma foetidissima\ see above, p. 284. 
I~I e lie hr y sum flic aide : dry grassy places. 
Veronica elliptica : a coast species. 
Piper excelsum : usually near the coast. 
Urtica australis. 
Pterostylis australis : cf. 11, p. 39. 
