XXIV 



and that the Deodar, as it retains its characters with the 

 least variation over the largest area, may be the original 

 type of a species embracing in its varieties the Cedars of 

 the Atlas and of Mount Lebanon, as well as those of the 

 Himalayan Mountains. That this is not mere hypothesis 

 may be shown, by an appeal to several New Zealand plants, 

 as, for instance, to those belonging to the genera Coprosma* 

 Celmisia,^ and Epilobium ; whilst it is illustrated in a man- 

 ner still more remarkable in the case of a Fern called Lo- 

 maria procera.\ In short, differences in habit, colour, hairi- 



* Coprosma is a genus of shrubs belonging to the Rubiacea, or Madder 

 tribe, confined to New Zealand, Tasmania, and the South Sea Islands, emitting 

 a very disagreeable smell when drying, and apt to vary very much in their 

 characters. Drawings of several species are given in the ' Flora Antarctica.' 



f Celmisia is a geaus of syngenesious plants, often very handsome, T)ut 

 varying much in their size and appearance. See 'Flora Antarctica' and 

 ' Flora Novse-Zelandise.' 



\ Lomaria procera is a common Fern in the Auckland Islands, in Australia, 

 and in Tasmania. Dr. Hooker distinguishes four principal varieties, but he 

 states that specimens of this fern vary so much, that no one would have taken 

 them all for the same species without examining the intermediate varieties. 

 The Fielding Herbarium at Oxford possesses two specimens which differ so 

 extremely in the shape of their pinnse, that it requires all one's faith in the 

 authority of the botanists who have investigated the subject to induce one to 

 regard them as merely varieties of the same species. Some of our British 

 Ferns indeed (Polypodium vulgare) vary very remarkably. 



