Notes. 
39 7 
of a stronger texture. When the fruit approaches the mature state, 
the inner part of the pericarp of these species begins to transform 
into a thin endocarp of a papery or cartilaginous nature, whilst the 
remainder dries up gradually. But whereas it hardly undergoes any 
particular transformation during that state in L. nicotianae/olia or in 
L. hypoleuca , a Hawaian species, the network of vascular bundles soon 
becomes very prominent in L. macrostachys and assumes the character 
of mechanical tissue which is formed by the elongated cells surrounding 
the vascular bundles becoming sclerenchymatous. These sclerenchy- 
matous masses are often confluent forming broader bundles, and they 
anastomose according to the disposition of the vascular bundles, thus 
forming a net-like woody skeleton. This skeleton immediately overlies 
the endocarp and is generally quite distinct, even before the latter has 
assumed its cartilaginous character. It is enclosed by a thick 
parenchymatous epicarp with a delicate epidermis. The epicarp dries 
up and contracts, but as the inner skeleton and the external ribs of 
the capsule are much stronger, it must break up itself. It becomes 
distinctly thinner over the interstices of the framework over which it 
is expanded. Minute cracks and holes appear, and soon a pore is 
formed showing the smooth margin of the corresponding mesh of the 
skeleton. As might be expected from this mode of perforation, which 
is not due to the presence of previously formed lines of weaker tissue 
like those along which the dehiscence of valves generally takes place, 
neither in the shape, size, number, nor position of the pores, nor their 
sequence is anything like regularity. In the material of the Kew 
Herbarium the number of pores is limited to few. But the fruits still 
bear the calyx-lobes and the decay of the epicarp is still in the first 
stage, whilst it is perfect in Zahlbruckner’s specimen. Dr. Zahl- 
bruckner denies the presence of any traces of insect-action or of 
wound-cork surrounding the pores. But both are present in the 
Kew specimens, and the only restriction I should like to make is as 
to their being caused by insects. I think it very probable, but I am 
not able to prove it from dry material. There are minute swellings 
on the surface, and a dissection shows that the epidermis had been 
injured and that wound-cork was formed subsequently from the 
nearest layers of the parenchyma. In other cases the wounds are 
larger, and their margins gape. They form a smooth rim which 
consists chiefly of a large-celled periderm. This periderm eventually 
permeates right through the epicarp, thus cutting out pores of a similar 
