NATURAL ORDERS. 117
tincnt of India, from Ceylon, Xcw Holland, and even from the Island of St. Vincent.
Jcrostichum stemaria of j\I. do Beauvois/ which hardly differs from A. alcicornc of New Holland, and of several of the islands of the Malayan Archipelago, was also ob- served ; and Acrosticlium aureum, which agrees with speci- mens from equinoctial America, was found growing in plenty among the mangroves near the mouth of the river.
I have formerly observed that the number of Filices, unhke that of the other Cryptogamous orders, (Lycopodineac excepted,) is greatest in the lower latitudes ; and, as I then supposed, near or somewhat beyond the tropics. The latter part of this statement, however, is not altogether correct ; the maximum of the order, both in absolute and relative number of species, being more probably within the tropics, though at considerable heights.
The degree of latitude alone being given, no judgment can be formed respecting the proportion of Tilices : for besides a temperature somewhat inferior, perhaps, to :ic-3 that of equinoctial countries of moderate elevation, a humid atmosphere and protection from the direct rays of the sun, seem to be requisite for their most abundant production.
AYhen all these conditions co-exist, their equinoctial proportion to Phecnogamous plants is probaljly about one to twenty, even on continents where the tracts most favourable to their production form only a small part, their number being increased according as such tracts constitute a more considerable portion of the surface.
Hence their maximum appears to exist in the high, and especially the wtU wooded, intratropical islands. Thus in Jamaica, where nearly two hundred species of Perns have been found, their proportion to Plia^nogamous plants is pro- bably about one to ten. In the Isles of France and Bourbon, from the facts stated by ]M. du Petit Thouars," they ai)i)ear to be about one to eight.
In Otaheite, according to Sir Joseph Banks's observations,
1^ more (VOicare 1, ;;. 2, ^. 2.
- Mtlanges de But. Observ, add. a M, de Lamarck, p. G, ct 38.
�� �
