TRANSACTIONS OF SECTION K. 995 
trace, under our very eyes, the first rise of anatomical characters which have else- 
where become of high taxonomic importance. A comparative study of the anatomy 
of any group of British plants, taking the same species growing under different 
conditions, would be sure to yield interesting results if any one had the patience to 
undertake it. 
Enough has been said to show that a given anatomical character may be of a 
high degree of constancy in one group while extremely variable in another, a fact 
which is already perfectly familiar as regards the ordinary morphological charac- 
ters. For example, nothing is more important in phanerogamic classification than 
the arrangement of the floral organs as shown in ground-plan or floral diagram. 
Yet Professor Trail’s observations, which he has been good enough to communicate 
to me, show that in one and the same species, or even individual, of Polygonum, 
almost every conceivable variation of the floral diagram may be found. 
There is, in fact, no ‘royal road’ to the estimation of the relative importance 
of characters ; the same character which is of the greatest value in one group may 
be trivial in another; and this holds good equally whether the character be drawn 
from the external morphology or from the internal structure. 
Our knowledge of the comparative anatomy of plants, from this point of view, 
is still very backward, and it is quite possible that the introduction of such charac- 
ters into the ordinary work of the Herbarium may be premature ; certainly it must 
be conducted with the greatest judgment and caution. We have not yet got our 
data, but every encouragement should be given to the collection of such data, so 
that our classification in the future may rest on the broad foundation of a com- 
parison of the entire structure of plants. 
In estimating the relative importance of characters of different kinds we must 
not forget that characters are often most constant when most adaptive. Thus, as 
Professor Trail informs me, the immense variability of the flowers of Polygonum 
goes together with their simple method of self-fertilisation. The exact arrange- 
ment is of little importance to the plant, and so variation goes on unchecked. In 
flowers with accurate adaptation to fertilisation by insects such variability is not 
found, for any change which would disturb the perfection of the mechanism is at 
once eliminated by natural selection. 
HisrToLoey. 
I propose to say but little on questions of minute histology, a subject which 
lies on the borderland between morphology and physiology, and which will be 
dealt with next Tuesday far more competently than I could hope to treat it. Last 
year my predecessor in the presidency of this Section spoke of a histological dis- 
covery (that of the nucleus, by Robert Brown) as ‘the most epoch-making of 
events’ in the modern history of botany. The histological questions before us at the 
present day may be of no less importance, but we cannot as yet see them in proper 
perspective. The centrosomes, those mysterious protoplasmic particles which have 
been supposed to preside over the division of the nucleus, and thus to determine 
the plane of segmentation, if really permanent organs of the cell, would have to 
rank as co-equal with the nucleus itself. If, on the other hand, as some think, 
they are not constant morphological entities, but at most temporary structures 
differentiated ad hoc, then we are brought face to face with the question whether 
the causes of nuclear division lie in the nucleus itself or in the surrounding 
protoplasm. 
Nothing can be more fascinating than such problems, and nothing more difficult. 
We have, at any rate, reason to congratulate ourselves that English botanists are 
no longer neglecting the study of the nucleus and its relation to the cell. Fora 
long time little was done in these subjects in our country, or at least little was 
published, and botanists were penealle content to take their information from 
abroad, not going beyond a mere verification of other men’s results. Now we 
have changed all that, as the communications to this Section sufficiently testify. 
Nothing is more remarkable in histology than the detailed agreement in the 
structure and behaviour of the nucleus in the higher plants and the: higher 
