1858.] BUNBURY FOSSIL LEATES, MADEIRA. 55 



appear to belong to a kind of laurel, — perhaps to more than one. 

 Heer has referred them to the Oreoda^lme fGetens, Nees {Laurus 

 foetens, Alton), or Til tree ; but the most perfect of the specimens in 

 the collection before me have, as it appears to me, a greater likeness 

 to the Laurus Canariensis, Webb. This, like the Til, is a very 

 abundant tree in the recent forests of Madeira. The most striking 

 character of the fossil leaves in question consists (as Heer has 

 already remarked) in the presence of certain little pits in the 

 axils of the lower primary veins, on the under side of the leaf, and 

 of corresponding protuberances or swelhngs on the upper side. In 

 this point they agree both with Oreodaphne foetens and with Laurus 

 Canariensis ; but the general shape and size of the leaves, and the 

 venation in general, in our best specimens, appear to me to corre- 

 spond best with the latter*. I am inclined to think, however, that 

 hoih occur among the S. Jorge plants. At the same time I must 

 say, that I have not seen one specimen sufficiently perfect to satisfy 

 me completely as to its specific characters ; those which are best 

 preserved as to outline do not exhibit the veining with sufficient 

 distinctness. The pits in the axils of the veins, though a striking 

 character, are not confined to the two species I have mentioned, for 

 I find similar pits in the leaves of at least two Brazilian species of 

 the same natural order. 



11. Oreodaphne foetens (Heer, No. 12. tab. 2. fig. 4-14). 



Some very ill-preserved specimens in the collection before me ap- 

 pear to agree with Heer's description and figures here quoted, especially 

 with his fig. 12. They differ from the leaves that I have referred to 

 Laurus Canariensis, principally in having the two lowest primary 

 veins nearly opposite to each other, and remarkably strong, so as 

 almost to give a triple-ribbed character to the leaf. In this they 

 agree with the Oreodaphne, and not with the L. Canariensis. The 

 network formed by the ultimate veins is somewhat different in the 

 two species, being more minute and uniform in the Laurus Canari- 

 ensis ; but none of the fossil specimens that I have seen are satis- 

 factory in this respect. The fossil flower represented by Heer (fig. 

 14) seems to agree with that of Oreodaphne ; I find nothing similar 

 among our specimens. 



12. Corylus australis (Heer, No. 10. tab. 2. fig. 1, 2). 

 Specimen very incomplete ; but its remarkable and well-preserved 



venation agrees exactly with that of Heer's plant. 



13. Salix ? ? 



A leaf, which, by its narrow linear lanceolate form, reminds us of 

 the Willows, and may possibly belong to that genus ; but it has the 

 appearance of a thicker and more coriaceous texture than is usual in 

 willows, and on the whole has, I think, more resemblance to some of 

 the narrow-leaved Myrtaceoe, such as Metrosideros angustifolia. The 

 veins are ill-preserved. It does not agree exactly with Heer's 

 Salix Lowei. 



* Laurus Canariensis, however, varies considerably in the shape of its leaves. 

 The pits in question are particularly conspicuous in tiae Azorean specimens 

 given me by Mr. Watson. 



