Linnaeus, in 1758, grouped the family into two divisions, by the form of the calyx, (five-fold or two-fold), and then went into the wildest confusion in distinction of species,—­sometimes by the form of corolla, sometimes by that of calyx, sometimes by that of the filaments, sometimes by that of the stigma, and sometimes by that of the seed.  As, for instance, thyme is to be identified by the calyx having hairs in its throat, dead nettle by having bristles in its mouth, lion’s tail by having bones in its anthers (antherae punctis osseis adspersae), and teucrium by having its upper lip cut in two!

14.  St. Hilaire, in 1805, divides again into four sections, but as three of these depend on form of corolla, and the fourth on abortion of stamens, the reader may conclude practically, that logical division of the family is impossible, and that all he can do, or that there is the smallest occasion for his doing, is first to understand the typical structure thoroughly, and then to know a certain number of forms accurately, grouping the others round them at convenient distances; and, finally, to attach to their known forms such simple names as may be utterable by children, and memorable by old people, with more ease and benefit than the ‘Galeopsis Eu-te-trahit,’ ‘Lamium Galeobdalon,’ or ’Scutellaria Galericulata,’and the like, of modern botany.  But to do this rightly, I must review and amplify some of my former classification, which it will be advisable to do in a separate chapter.

* * * * *

CHAPTER VI.

MONACHA.

1.  It is not a little vexing to me, in looking over the very little I have got done of my planned Systema Proserpinae, to discover a grave mistake in the specifications of Veronica.  It is Veronica chamaedrys, not officinalis, which is our proper English Speedwell, and Welsh Fluellen; and all the eighth paragraph, p. 74, properly applies to that.  Veronica officinalis is an extremely small flower rising on vertical stems out of recumbent leaves; and the drawing of it in the Flora Danica, which I mistook for a stunted northern state, is quite true of the English species,[32] except that it does not express the recumbent action of the leaves.  The proper representation of ground-leafage has never yet been attempted in any botanical work whatever, and as, in recumbent plants, their grouping and action can only be seen from above, the plates of them should always have a dark and rugged background, not only to indicate the position of the eye, but to relieve the forms of the leaves as they were intended to be shown.  I will try to give some examples in the course of this year.