846 
MR, W. GARDINER ON THE CONTINUITY OF THE 
walls as well. I have observed that this happens in Kentia Belmoriana, Kentia costcita, 
Lodoicea, BentincJcia, and Asperulct (see figs. 16, 17, 19, 31); but I am led to believe 
that such union is of much more general occurrence. In all the foregoing examples, 
the threads running through the walls are more especially obvious in the cells just 
below the surface, and gradually become less and less visible, as one approaches the 
central tissue of the seed. This appears to me to be simply an arrangement for 
insuring that every facility should be given for the passage of nutritive material from 
without inwards, and also that it should have opened to it as many channels as 
possible. It is obvious, for instance, that the amount of plastic formative substance 
required for building up such a tremendous endosperm as that of Lodoicea must be 
very considerable, and even supposing its growth to be slow, the drain on the 
nutritive material must be large, and the rate of its flow must be very great. 
Consequently the increased facilities for easy transmission must be of great 
advantage. And not only in the development of the endosperm, but also upon 
germination, is this structure of great use to the plant, for at that period the outer 
cell-layers will have become very dry, and consequently the difficulty of their being- 
broken down by the absorbent foot of the cotyledon will be increased. But at the 
same time, owing to the greater development of a system of channels in them, they 
are more easily permeated and wetted by the cell sap holding in solution the 
ferment which will bring about their final disorganisation. 
The form presented by the aggregate of threads traversing the pit membrane is 
usually that of the well-known basket or barrel-shape which is met with in connexion 
with nuclear division. In many instances, and especially in BentincJcia, the shapely 
sweep of the curving threads, and the graceful arrangement of the whole thread- 
complex is extremely striking and beautiful (Plate 69, fig. 16). In other cases the 
bending of the curve is not so marked, and in very thin pit membranes— e.g., Syne- 
clianthus, Livi.stona, &c.—the threads appear to be altogether straight. 
In the instances where the threads go through the cell-wall their direction is 
seldom straight, but usually bent, and resembling in arrangement the appearance 
presented by the achromatic fibres during free-cell-formation. 
Every variation occurs both as regards the size of the cells, the distribution of the 
pits, and the number and thickness of the threads. Thus, whereas the cells of 
Caryota urens and Lodoicea (Plate 69, fig. 19) are large, those of Thrinax and 
Geonoma are small. In such endosperms as Manicarea and CJiamceclorea the pits 
are very numerous, while in Washingtona but few are present. In Calyptronoma the 
threads are few and somewhat stout, while in Oncosperma they are very numerous 
and fine. In BentincJcia and HeterospatJie they are also many in number. The 
threads are made very conspicuous by staining with iodine and treatment with Clilor. 
Zinc lod., for the latter reagent appears to cause a decided precipitation of iodine upon 
them as well as upon the general protoplasm which is accompanied by an increase in 
