and never define a 'horn' all the while!

Viola Cornuta, however, let it be; for the name does mean something, and is not false Latin. But whether violet or pansy, I must look farther to find out.

23. I take the Flora Danica, in which I at least am sure of finding whatever is done at all, done as well as honesty and care can; and look what species of violets it gives.

Nine, in the first ten volumes of it; four in their modern sequel (that I know of,—I have had no time to examine the last issues). Namely, in alphabetical order, with their present Latin, or tentative Latin, names; and in plain English, the senses intended by the hapless scientific people, in such their tentative Latin:—

I next run down this list, noting what names we can keep, and what we can't; and what aren't worth keeping, if we could: passing over the varieties, however, for the present, wholly.

(1) Arvensis. Field-violet. Good.

(2) Biflora. A good epithet, but in false Latin. It is to be our Viola aurea, golden pansy.

(3) Canina. Dog. Not pretty, but intelligible, and by common use now classical. Must stay.

(4) Hirta. Late Latin slang for hirsuta, and always used of nasty places or nasty people; it shall not stay. The species shall be our Viola Seclusa,—Monk's violet—meaning the kind of monk who leads a rough life like Elijah's, or the Baptist's,

or Esau's—in another kind. This violet is one of the loveliest that grows.

(5) Mirabilis. Stays so; marvellous enough, truly: not more so than all violets; but I am very glad to hear of scientific people capable of admiring anything.

(6) Montana. Stays so.

(7) Odorata. Not distinctive;—nearly classical, however. It is to be our Viola Regina, else I should not have altered it.

(8) Palustris. Stays so.

(9) Tricolor. True, but intolerable. The flower is the queen of the true pansies: to be our Viola Psyche.

(10) Elatior. Only a variety of our already accepted Cornuta.

(11) The last is, I believe, also only a variety of Palustris. Its leaves, I am informed in the text, are either "pubescent-reticulate-venose-subreniform," or "lato-cordate-repando-crenate;" and its stipules are "ovate-acuminate-fimbrio-denticulate." I do not wish to pursue the inquiry farther.

24. These ten species will include, noting here and there a local variety, all the forms which are familiar to us in Northern Europe, except only two;—these, as it singularly chances, being the Viola Alpium, noblest of all the wild pansies in the world, so far as I have seen or heard of them,—of which, consequently, I find no picture,

nor notice, in any botanical work whatsoever; and the other, the rock-violet of our own Yorkshire hills.

We have therefore, ourselves, finally then, twelve following species to study. I give them now all in their accepted names and proper order,—the reasons for occasional difference between the Latin and English name will be presently given.

25. We will try, presently, what is to be found out of useful, or pretty, concerning all these twelve violets; but must first find out how we are to know which are violets indeed, and which, pansies.

Yesterday, after finishing my list, I went out again to examine Viola Cornuta a little closer, and pulled up a full grip of it by the roots, and put it in water in a wash-hand basin, which it filled like a truss of green hay.

Pulling out two or three separate plants, I find each to consist mainly of a jointed stalk of a kind I have not yet described,—roughly, some two feet long altogether; (accurately, one 1 ft. 10½ in.; another, 1 ft. 10 in.; another, 1 ft. 9