Page:A Voyage to Terra Australis Volume 2.djvu/598

584 fifth volume of Willdenow's edition of the Species Plantarum. In their geographical distribution Ferns differ from all the other orders of cryptogamous plants, their maximum being in the lower latitudes, probably near, or very little beyond the tropics. Thus Norfolk Island, situated in 29° S. lat. and only a few leagues in circumference, produces as many species of the order as are described in Dr. Smith's Flora Britannica.

But as shade and moisture are essential conditions to the vegetation of the greater part of Ferns, few species only have been observed in those parts of æquinoctial New Holland, hitherto examined. The number of species already found, however, in the different regions of Terra Australis exceeds 100, of which, nearly one-fourth are also natives of other countries.

Among the Australian Ferns there is no genus absolutely confined to that country, except Platyzoma, but this, perhaps, ought not to be separated from Gleichenia.

Only two arborescent Ferns have been observed in Terra Australis, one in the colony of Port Jackson, the second, Dicksonia antarctica, is frequent in Van Diemen's Island, at the southern extremity of which its trunk is not unfrequently from 12 to 16 feet in height. An arborescent species of the same genus was found by Forster, in New Zealand, at Dusky Bay, in nearly 46° S. the highest latitude in which tree ferns have yet been observed. It is remarkable that, although they have so considerable a range in the southern hemisphere, no tree fern has been found beyond the northern tropic: a distribution in the two hemispheres somewhat similar to this has been already noticed respecting the Orchideæ that are parasitical on trees.

I have formerly, in treating of the New Holland Asplenia, observed that Cœnopteris does not differ from them in the relation its involucra have to the axis of the frond or pinna, but merely in having the ultimate pinna more deeply divided with one, or, at most, two involucra on each segment, towards the margins of which they must necessarily open: hence, the characters of both genera not unfrequently occur in the same frond, and are even exhibited by the same involucrum when it happens to extend below the origin of the segment.

I have observed also, in the same place, that in Asplenium when the