ee le 
TRANSACTIONS OF SECTION D. 537 
The author sums up his conclusions as follows :— 
1. That protoplasmic threads connecting neighbouring protoplasts are present in 
such widely different and diffused structures as sieve-tubes, cortical parenchyma, 
leaf-pulvinus, pith of resting leaf-bud, and endosperm of seeds. 
2. That in the contraction of the protoplast in natural plasmolysis these threads 
would normally remain unbroken. 
3. That they may serve to transmit impulses from one cell to another, 
acting in this way somewhat like a nervous system. 
4, That besides the perforating threads, equally widely spread and much more 
numerous, are threads which attach the protoplast to the cell wall, whether at the 
base of pits or otherwise ; and that these threads are often opposite to each other. 
5. That the closing membrane separating two threads often shows differentiation 
which suggests permeability, if not ‘sieve perforation.’ 
6. That in contraction of the protoplast in natural plasmolysis these threads 
also would be normally unbroken. 
7. That these threads may when in extension act upon the cell-wall and put it 
in a state of slight positive tension. 
8. That the presence of minute perforations communicating from cavity to 
cavity of living cells, wovdd not, and if communicating with the intercellular spaces 
need not, be a hindrance to the turgipotence of the cells. 
5. On some Cell Contents. By MarsHatt Warp. 
The author has for some time past been engaged in researches among the 
fungi—particularly those which attack living plants, and his attention was 
necessarily directed to tke cell-contents of the host plants: among others, the cells 
of Coffea, Cinchona, Pavetta, and Canthium, and one or two cryptogams have 
received special attention. The present paper refers particularly to one class of 
bodies found in the cells of the cultivated species of Coffea—C. arabica, C. 
liberica, &e. 
The structure and chemical character of the endosperm have been carefully 
investigated, and the author gives details of the germination. 
Certain fatty bodies, mixed with proteids in the endosperm, are traced into the 
embryo and seedling, and their reactions and changes are noticed. 
In the leaves, cortex, and other soft parts of the mature plant, are found ‘ fat 
bodies’ under circumstances which compel the author to conclude that they are 
the results of constructive activity, and not products of destructive metabolism. 
These ‘fat bodies’ consist of varying mixtures of fats and other substances— 
probably in part proteid—and show considerable similarity to the fatty masses in 
the endosperm. 
Details are given of their reactions and changes, and the author believes that 
they represent temporary stores, to be worked up further in the construction of 
higher bodies. They may be, in fact, fusions of fatty matters in various stages 
of transition, carbohydrates, and salts of sulphur, nitrogen, &c., derived from the 
soil: if so, the author points out that it is not improbable that the earlier 
constructive acts in the formation of proteids may be taking place here. 
6. On the Nectar Gland of Reseda. 
By Professor ALEXANDER S. Wixson, M.A., B.Sc. 
The flowers of this genus, of which the garden mignonette is the best known 
example, possess in addition to the ordinary floral organs an extra structure named 
adisk. This disk, which is an expansion of the top of the flower-stalk, is hollow on 
its upper surface, its shape resembling one-half of a bivalve shell. In this shell- 
shaped cavity of the disk the honey is secreted, and above it is protected by the 
three upper petals, the claws of which are flattened and closely overlap, forming a 
lid which completely closes the nectar-holder. The honey of Reseda is thus 
contained within a closed box, the lid of which must be prized up before it can be 
