is above described with attention, vol. i., p. 75, as an example of precision in flower-form, we may as well retain it in our list here. It will be therefore our twenty-first variety,—it is Loudon's fifty-ninth and last. He translates 'polita' simply 'polished,' which is nonsense. I can think of nothing to call it but 'dainty,' and will leave it at present unchristened.
5. Lastly. I can't think why I omitted V. Humifusa, S. 979, which seems to be quite one of the most beautiful of the family—a mountain flower also, and one which I ought to find here; but hitherto I know only among the mantlings of the ground, V. thymifolia and officinalis. All these, however, agree in the extreme prettiness and grace of their crowded leafage,—the officinalis, of which the leaves are shown much too coarsely serrated in S. 984, forming carpets of finished embroidery which I have never yet rightly examined, because I mistook them for St. John's wort. They are of a beautiful pointed oval form, serrated so finely that they seem smooth in distant effect, and covered with equally invisible hairs, which seem to collect towards the edge in the variety Hirsuta, S. 985.
For the present, I should like the reader to group the three flowers, S. 979, 984, 985, under the general name of Humifusa, and to distinguish them by a third epithet, which I allow myself when in difficulties, thus:
V. Humifusa, cærulea, the beautiful blue one, which resembles Spicata.
V. Humifusa, officinalis, and,
V. Humifusa, hirsuta: the last seems to me extremely interesting, and I hope to find it and study it carefully.
By this arrangement we shall have only twenty-one species to remember: the one which chiefly decorates the ground again dividing into the above three.
6. These matters being set right, I pass to the business in hand, which is to define as far as possible the subtle relations between the Veronicas and Draconidæ, and again between these and the tribe at present called labiate. In my classification above, vol. i, p. 200, the Draconidæ include the Nightshades; but this was an oversight. Atropa belongs properly to the following class, Moiridæ; and my Draconids are intended to include only the two great families of Personate and Ringent flowers, which in some degree resemble the head of an animal: the representative one being what we call 'snapdragon,' but the French, careless of its snapping power, 'calf's muzzle'—"Muflier, muflande, or muffle de Veau."—Rousseau, 'Lettres,' p. 19.
7. As I examine his careful and sensible plates of it, I chance also on a bit of his text, which, extremely wise and generally useful, I translate forthwith:—
"I understand, my dear, that one is vexed to take so much trouble without learning the names of the plants one examines; but I confess to you in good faith that it never entered into my plan to spare you this little
chagrin. One pretends that Botany is nothing but a science of words, which only exercises the memory, and only teaches how to give plants names. For me, I know no rational study which is only a science of words: and to which of the two, I pray you, shall I grant the name of botanist,—to him who knows how to spit out a name or a phrase at the sight of a plant, without knowing anything of its structure, or to him who, knowing that structure very well, is ignorant nevertheless of the very arbitrary name that one gives to the plant in such and such a country? If we only gave to your children an amusing occupation, we should miss the best half of our purpose, which is, in amusing them, to exercise their intelligence and accustom them to attention. Before teaching them to name what they see, let us begin by teaching them to see it. That science, forgotten in all educations, ought to form the most important part of theirs. I can never repeat it often enough—teach them never to be satisfied with words, ('se payer de mots') and to hold themselves as knowing nothing of what has reached no farther than their memories."