TRANSACTIONS OF SECTION K. 993 
characters, there is a great difference in the morphologist’s and the physiologist’s 
way of looking at them. ‘The physiologist is interested in the question how 
organs work ; the morphologist asks, what is their history ? 
The morphologist may well feel discouraged at the vastness of the work before 
him. The origin of the great groups of plants is perhaps, after all, an insoluble 
problem, for the question is not accessible either to observation or experiment. 
All that we can directly observe or experiment upon is the occurrence of varia- 
tions—perhaps the most important line of research in biology, for it was the study 
of variation that led Darwin and Wallace to their grand generalisation. Many 
observers are working to-day in the spirit of the great masters, and it is certain 
that their work will be fruitful in results. It is evident, however, that such 
investigations can at most only throw a side light on the historical question of the 
origin of the existing orders and classes of living things. The morphologist has 
to attack such questions by other methods of research. 
The embryological method has so far scarcely received justice from botanists. 
A great deal of what is called embryology in botany is not embryology at all, 
but relates to pre-fertilisation changes. Of real embryology—that is to say, the 
development of the young plant from the fertilised ovum—there is much less than 
we might expect. Thus no comparative investigation of the embryology of either 
Dicotyledons or Monocotyledons has ever been carried out, our knowledge being 
entirely based on a few isolated examples. 
In the cases which have been investigated perhaps excessive attention has been 
devoted to the first divisions of the ovum, the importance of which, as Sachs long 
ago showed, has been overrated, while the later stages, when the differentiation 
of organs and tissues is actually in progress, have been comparatively neglected. 
The law of recapitulation (or repetition of phylogeny in ontogeny) has been 
very inadequately tested in the vegetable kingdom. Whatever its value may be, 
it is certainly desirable that the development of plants as well as animals should 
be considered from this point of view ; and this has so far been done in but very 
few cases. M. Massart, of Brussels, has made some investigations with this object 
on the development of seedlings and of individual leaves. He is led to the con- 
clusion that examples of recapitulation are rare among plants.’ 
So far, at least, embryological research has only yielded certain proof of re- 
capitulation in a few cases, as in the well-known example of the phyllode-bearing 
acacias, in which the first leaves of the seedling are normal, while the later formed 
ones gradually assume the reduced phyllode form. 
A less familiar example is afforded by Gunnera. Here, as is well known, the 
mature stem has a structure totally different from that of ordinary Dicotyledons, 
and much resembling that characteristic of most Ferns. In most species of 
Gunnera there are a number of distinct vascular cylinders in the stem, instead of 
one only, and there is never the slightest trace, so far as the adult plant is con- 
cerned, of the growth by means of cambium, which is otherwise so general in the 
class. The seedling stem, however, is not only monostelic below the cotyledons, 
but in this region, though nowhere else, shows distinct secondary growth. Thus, 
if we were in any doubt as to the general affinities of Gwnnera, owing to its 
extraordinary mature structure, we should at once be put on the right track by the 
study of the embryonic stem, which alone retains the characteristic dicotyledonous 
mode of growth. 
It is only in a few cases, however, and for narrow ranges of affinity, that the 
doctrine of recapitulation has at present helped in the determination of relationships 
among plants. Beyond this, conclusions based on embryology alone tend to 
become merely conjectural and subjective. In fact, all comparative work, in so 
far as it is limited to plants now living, suffers under the same weakness 
that it can never yield certain results, for the question whether given characters 
are relatively primitive or recently acquired is one upon which each naturalist is 
left to form his own opinion, as the origin of the characters cannot be observed. 
La Récapitulation et l’Innovation en Embryologie Végétale,’ Bull. de la Soc. 
roy. de Bot. de Belgique, vol, xxxiii., 1894, 
