leaflets there were in the leaf, three, five, seven, nine, or occasionally
six or eight.

[Footnote 1: Bud-scales are modified leaves and their arrangement is
therefore the same as the leaves. This is not mentioned in the study of
the Horsechestnut bud, because it cannot be proved to the pupils, but the
transition is explained in connection with Lilac, where it may be clearly
seen. The scales of the bud of Horsechestnut are considered to be
homologous with petioles, by analogy with other members of the same
family. In the Sweet Buckeye a series can be made, exhibiting the gradual
change from a scale to a compound leaf. See the Botanical Text-Book, Part
I, Structural Botany. By Asa Gray. Ivison, Blakeman, Taylor and Co., New
York, 1879. Plate 233, p. 116.]

[Illustration: FIG. 13.--Horsechestnut. I. Branch in winter state: _a_,
leaf-scars; _b_, bud-scars; _c_, flower-scars. 2. An expanding leaf-bud.
3. Same, more advanced.]

_The Bud Scale-Scars_. These are rings left by the scales of the bud and
may be seen in many branches. They are well seen in Horsechestnut. If the
pupils have failed to observe that these rings show the position of former
buds and mark the growth of successive years, this point must be brought
out by skilful questioning. There is a difference in the color of the more
recent shoots, and a pupil, when asked how much of his branch grew the
preceding season, will be able to answer by observing the change in color.
Make him see that this change corresponds with the rings, and he will
understand how to tell every year's growth. Then ask what would make the
rings in a branch produced from one of his buds, and he can hardly fail to
see that the scales would make them. When the scholars understand that the
rings mark the year's growth, they can count them and ascertain the age
of each branch. The same should be done with each side-shoot. Usually the
numbers will be found to agree; that is, all the buds will have the
same number of rings between them and the cut end of the branch, but
occasionally a bud will remain latent for one or several seasons and then
begin its growth, in which case the numbers will not agree; the difference
will be the number of years it remained latent. There are always many buds
that are not developed. "The undeveloped buds do not necessarily perish,
but are ready to be called into action in case the others are checked.
When the stronger buds are destroyed, some that would else remai