and the Pyrenees^ in 1825. 273 



^' Through all the districts I have passed during these three 

 days, I observed that the prevailing large plants on the waste 

 lands are Genista scorpius. Cist us monspeliensis, and Lavandula 

 spica, interspersed with Quercus ilex and coccifera, Rosmarinus 

 officinalis (the Rosemary), Buxus sempervirens (the Box- tree), 

 &c. Of these, the Genista scorpius holds the place of the Ulex 

 £uropeeus or Furze in Scotland : the Lavandula spica * they 

 have in place of our Calluna erica, or heather ; and as for the 

 Quercus ilex and coccifera, they would, if not cropped by the 



» sheep, cover the grounds like the Quercus sessiliflorus in the 

 Highlands of Scotland. Upon the whole, the country here is 

 far from beautiful : the whole, with the exception of a few scat- 

 tered valleys, consisting of waste lands or Garriques. These 



\ wastes consist of little but small sharp-cornered hard stones, 

 of a grey colour, which in some places ages scarcely moulder 

 down : the consequence is, that there is but little soil, and 

 that little is immediately seized upon by the wild plants. 

 Nor is it possible to cultivate these garriques ; remove the 

 surface stones, and you only find more ; besides, the heavy 

 October rains wash down the finest mould into the valleys, 

 leaving the hilly parts absolutely destitute of covering. Cale- 

 donians hills are celebrated for being bare and barren ; but they 

 must give up the point when Languedoc contests it. On ac- 



• The plant found at Montpellier, and throughout Provence, is the true 

 L. spica; that usually cultivated in our gardens is a more mountainous species, 

 being the L. vera of De CandoUe. Between these two and the L. pyrenaica 

 there exists great confusion among authors. The chief character pointed 

 out by De CandoUe is the shape of the bractese being linear in L. spica, and 

 broadly cordato-acuminate in the other two ; and even in these, the shape is 

 not precisely the same, though not so strikingly different as to afford a good 

 specific character. De CandoUe, however, founds upon it. One of the nerves 

 of the calyx in all the species expands at the apex into a smaU foliaceous ap- 

 pendage, which closes the orifice of the calyx before the appearance of the co- 

 rolla. The shape of that part, although the character be extremely minute, 

 affords, as my friend the Baron Gingins de Lassarey at Berne has clearly 

 shewn me, the best specific difference. Baron Gingins, the celebrated au- 

 thor of the article ViolariecB in De CandoUe's Prodromus, is about to publish a 

 monograph on the Lavandula. Among other remarkable discoveries, he has 

 found that the simple and pinnated leaved species present two very different 

 structures in the embryo of the seed, which promises fair not only to separate 

 the latter generically from the former, but, should a natural classification of 

 the LabiatcB demand it, to remove them to a different part of the order. 



JULY OCTOBER 1826. S 



