110 REPORT—1868. 
new stratum of wood being formed by the cambium outside the old wood, and a 
new stratum of bark on the inside of the full-grown bark. From this ordinary 
mode there are occasional deviations. In a former paper I directed the attention 
of botanists to one exception in the case of Menispermaceze, where at the end of 
every three or four years the cambium becomes inactive, and a new one is formed 
outside of the bast. 
Having been for some time engaged in the systematic study of the difficult tribe 
of Sapindacez (the ‘ crux Botanicorum”), I have had abundant opportunities of 
examining the structural peculiarities of this most interesting tribe of plants, the 
small branches of herbarium specimens exhibiting the same peculiarities which 
strike observers with so much surprise in large trunks. 
I do not propose at present to explain in detail the development of these irregu- 
larities in Sapindaceze, nor to trace their relations to other families, such as 
Bignoniaceze, Malpighiaceze, &c. My object is to state as briefly as possible the 
most remarkable modifications of wood structure which I have observed. To enter 
into details, it would be necessary to have the specimens themselves at hand, 
but it would not have been possible to bring with me the numerous materials 
which I have had from the Berlin, Vienna, Copenhagen, St. Petersburgh, and other 
Herbaria. 
The two genera in which I have observed anomalous stem-structure are Serjania 
and Paullinia, and in both normal stems also occur. In Pavllinia I have noticed 
only one type of irregularity with some slight modifications, and as the same type 
with nearly the same modifications is found in Serjania, I shall confine my remarks 
to the latter genus. 
If we examine transverse sections of young branches (one to three years old) of 
Serjania, we seldom find a circular wood. In most species it is angular, trigonal 
or pentagonal. In some species the angles are very prominent, so that the wood 
is deeply furrowed, or, what is yet more striking, the angles become detached from 
the centre, so that the wood is compound, 
Different kinds of compound wood occur in different species. There may be as 
many peripheral detached parts as there are angles, or there may be fewer, some- 
times only one; or ayain, there may be more, often as many as eight or ten nearly 
touching one another, so as to form a ring round the central wood. The peripheral 
parts may be either cylindrical with the stem indented, or the parts may be flattened 
in different degrees, and in the latter case the stem is smooth, without any indica- 
tion of the internal irregularity. 
Besides all these modes of irregularity, there is yet another which cannot be 
brought into connexion with them, and which I have not seen anywhere described. 
It consists in the wood being split by radial divisions into five nearly equal portions. 
This I call divided wood. 
We can thus distinguish in different species of Serjania, round wood, poly- 
gonal wood, furrowed wood, compound wood, with as many or more parts 
than there are angles, and finally divided wood. Now I have found that in each 
species of Serjania and Paullinia the form of stem is constant, provided the obser- 
vation be made in the right place, namely about the middle of the branch, and not 
at the lower part near its origin, where the different parts of the compound and 
divided wood are commonly united into one single wood. The section must also 
not be made immediately below a leaf; for there the number of peripheral parts is 
frequently diminished by their union with the central wood. The neglect of these 
precautions has led to the belief by former observers that there is no constancy in 
each species. There is, however, certainly such constancy, and as it is now for the 
first time pointed out, I desire particularly to direct attention to it, as well as to the 
further fact, that the structure of the wood corresponds closely with the specific 
characters derived from the flower and fruit, so that groups of species formed from 
the wood-structure will be nearly identical with those based on flower and fruit- 
structure, and may therefore be considered as quite natural. 
It would lead me too far were I to attempt to lay before you an abstract of these 
natural groups thus formed. They will appear in a general work which I am 
publishing on Sapindaceze, where they willbe accompanied by the necessary plates. 
I may mention here, that the grouping of the species of Pauilinia and Serjania 
