289 
and Histogeny of Strychnos . 
new elements can only find room by the crushing and oblitera- 
tion of the old. The nearer the island lies to the centre of 
the stem (i. e. the older it is) the greater of course is the 
amount of obliteration. In the oldest observed, the mass of 
crushed elements, which as we have seen always lies on the 
outer side of the island, may occupy almost half its area (see 
Fig. 1 1). Towards the inner side, on the other hand, a distinct 
cambial layer, with thin tangential cell-walls, is always found. 
As we have already seen, it is a constant rule that the radial 
arrangement of the phloem-elements is most regular in the 
inner part of the island, and that the radial series can be 
traced continuously inwards through the cambium into the 
wood. Towards the outside the radial rows, even when not 
wholly erased by obliteration, can seldom be traced at all, and 
never with any regularity. That any signs of radial continuity 
can ever be seen on the outside is due to the fact that the 
complementary cambium (in this species at any rate) arises by 
the division of cells which were themselves of cambial origin. 
Some account must be given of the position of the phloem- 
islands in the wood ; and first, as regards their relation to the 
medullary rays. A small island often occupies the space be- 
tween two rays. In other (rarer) cases, a medullary ray can 
be clearly traced straight through the island. In one case a 
large ray, three cells thick, was observed to pass through the 
middle of an island, its cells being thin-walled in this part of 
its course. At the outer margin of the phloem the cells of the 
ray are not obviously distinct from the parenchyma so often 
found at the outside of an island, but in the external wood the 
ray goes on again as before. These facts are explained by 
the development. The complementary cambium has simply 
formed new ray-cells which approximately fit on to those 
which the normal cambium had begun to form before its dis- 
location. Not uncommonly however a broad secondary ray 
starts from the outer margin of a phloem-island. In other 
cases a small ray can be traced to the inner margin of an 
island and through the cambial layer into its tissue, but it 
does not reappear on the outside. The explanation of all 
