54 ° Brebner. — On the Anatomy of 
(1) The sclerenchyma. The petiolar sclerenchyma in most 
of the Marattiaceae is a simple continuous band of thick- 
walled elements a little below the epidermis, Fig. 13, set., 
but in Angiopteris its inner margin is sculptured owing to 
circular groups of the sclerenchyma projecting inwards. 
There is no such sculpturing in any of the five species of 
Marattia studied, so that as far as they are concerned the 
genus is distinguished from Angiopteris by this histological 
feature. In Danaea elliptica , however, an almost identical 
state of matters is met with in the fertile, but not the barren 
fronds. This is very curious and difficult to explain. It is 
evident therefore that the arrangement of the elements of the 
sclerenchyma may have some diagnostic value. This differ- 
ence in the sclerenchyma of the fertile and barren fronds of 
Danaea elliptica is one example of the many pitfalls and 
snares which beset the path of the palaeophytologist. 
(2) Mucilage-canals. Although in general there is no 
specially characteristic arrangement of the mucilage-canals, 
all the same, one species of Marattia showed a distinctive 
peculiarity in this respect, viz. M. cicutaefolia. In this species 
only was found in the pulvinus a ring of mucilage-canals 
lying between the collenchyma and the epidermis. 
(3) Tannin-sacs. The number and arrangement of these 
structures vary a good deal in the different species, and it 
was at first thought they would be useful as systematic 
guides, but revision of the matter showed that there is not 
sufficient constancy and definiteness in arrangement to be 
of real value. However, careful plotting of their position and 
comparison of the arrangement in the different species might 
occasionally aid specific discrimination. Tannin-sacs, for 
instance, are abundantly present round the meristeles in the 
pulvinus of M. cicutaefolia and M. attenuata , and they are 
practically absent in the pulvinus of M. fraxinea. The 
section from which Fig. 11 was drawn had only two quite 
indefinitely placed tannin-sacs. 
No attempt was made to test the distribution of crystals 
from this point of view, although in the hands of Radlkofer 
