ON AFHYLLOSTACHYS. 225 
in my c Bernstein Flora, 5 and as has since been confirmed by linger 
and Hartig. 
From all this it is evident that new species appeared and disap- 
peared at all times, and that at no times were all the species of plant 
simultaneously created or simultaneously swept away. 
2. A few Orders and families attain, on their first appearance, b 
high degree of development, and retain this down to the present time. 
This applies to the oldest family, the Alga, for I have discovered 
Flondea in the Silurian formation ; and also to a somewhat younger 
Order, that of the Ferns, which even in the first terrestrial flora, at- 
tained a high development, and has retained it through all the forma- 
tions up to the present time, without ever having experienced any 
transmutation or a period of evolution. Other Orders first appear in 
some isolated divisions or families, as, for instance, the Cbnifera, which 
began with the Abittinerz, and gradually became more complex, but 
as early as in the Palaeozoic period appeared in such diversity of 
wm 3 and consequently high internal structure, as in no subsequent 
period. This is most important, as, for instance, one of their pecu- 
liarities is their having compound medullary rays (instead of the simple 
forms which now prevail in the whole family), reminding us markedly 
°1 the Dicotyledonous type which did not appear before the Cretaceous 
Period. But this high degree of development is confined to the 
Abietmea ; the Con if era and Cupremnea of the Permian, and the 
v&nea and Gnetacea of the Tertiary period correspond in all pecu- 
liarities with those of the present time. An equally high develop- 
ment of internal structure was attained by the Cycadece (closely allied 
^ on \fera 9 and also belonging to Gymnosperms), as early as the 
wniau formation (/. e. towards the close of the Palaeozoic period), in 
the curious trunks of Medullow stellata, Cotta. The structure and 
1 * 
sposition of the woody cylinder and its medullary rays correspond 
generally with those of Encephalartos, a Cycad of the present flora, but 
w *th this great difference, that there are in the pith not only isolated 
jocular bundles as in Eucephalartos, but a complete woody cylinder, 
•wing the structure of the principal cylinder enclosing them. These 
Antral cylinders repeat the structure of the whole trunk, and the 
r ganism has attained, on this account, a higher stage of development 
um is observed in any Cycad before or since that time. Moreover 
118 stl 'ucture is unique in the whole vegetable kingdom, as recent exa- 
