2 
completion within a practicable time. It is like the bold 
treatment of the first cast of a monumental work and, rightly 
and fairly understood, does not detract from the value of the 
work or the genius of the author. On the other hand, it challenges 
by its very nature the criticism of the specialist who is called 
upon to supply the more minute details with a view towards 
a final recognition of the taxonomic structure of the group- 
Neither the material available, nor the insight which we possess 
into the real value of the differences which present themselves 
in specimens mostly collected in a haphazard way, badly 
annotated, if at all, and frequently imperfect in some respects, 
warrants a claim to that finality having been reached. It is 
evident that nothing but painstaking and comprehensive analysis 
and the avoidance of unsupported generalisation can lead to 
the desired goal. 
It is hoped that it is in this sense that the Australian botanists, 
working at the sources of our knowledge of those grasses, will 
receive this attempt at unravelling the threads of an unusually 
intricate taxonomic problem. 
In explanation of the procedure adopted, I may add that: 
the grouping of the species as it appears in the Key that precedes - 
the Enumeration of the Species was obtained almost entirely 
from the comparison of external morphological characters such 
as a careful and catholic analysis revealed. Groups of wider 
and narrower extent presented themselves, and those which 
could not be further analysed without drifting away from 
apparently stable and therefore reliable characters were accepted 
as species. Some of the wider groups were obvious enough to 
be recognised at a glance, others less definitely circumscribed, 
suggested themselves in the first place through their more 
marked members which were found to form nuclei round which 
others clustered more or less evidently. Out of this combination 
of analysis and synthetic perception the Key grew almost 
automatically. The differentiation thus revealed of the whole 
plexus of forms under consideration rests, therefore, on a purely: 
empirical basis such as might be expected to reflect natural 
relationships. 
It was not until after the drafting of the Key was completed 
that I suggested that the result obtained should be tested by 
reference to the anatomical characters of the leaf-blades, partly 
to see how they compared with the grouping as it manifested 
itself through the external characters, and partly to gain addi- 
tional data which might assist in the recognition of the species. 
e result of Miss Hughes’ work in that direction is embodied 
in the text figures on pp. 19 and 29. The common experience 
that the variation of the anatomical structure of the leaf-blade 
of the grasses does not generally run parallel with the characters 
from which we infer their phylogenetic relationships, although 
they add frequently to the distinctiveness of the species, also 
holds good in the present case. Only one group (Sclerophylle) 
