o 
THE MAGNOLIA FAMILY 359 
in figures 159,2 and 271. Somewhat similar expansions serv- 
ing for protection occur at the base of marsh-marigold leaves; 
but these, although suggesting stipules, are not regarded as 
sufficiently developed to deserve the name. The leaves of 
star-anise, as of the crowfoot family, are exstipulate,! that is, 
without stipules. Finally, as regards their habit,? or general ap- 
pearance, the tulip-tree is, as itsname implies, a tree, while the 
species of magnolia and star-anise are either trees or shrubs. 
The result of our examination thus far is to show that 
star-anise in several particulars forms a good link connecting 
the tulip-tree with members of the crowfoot family, and we 
have not yet found a single feature which will serve to dis- 
tinguish all of the magnolia family from all of the crowfoot 
family. 
This resemblance will appear still more plainly if we express in 
formulas the facts observable in our examples. Let us indicate the 
presence of stipules by an inverted dagger sign, |; a wing on the 
pericarp by an inverted interrogation mark, Z; and dorsal dehiscence 
by >. We may then write our formulas of Magnolia, Illicitum, and 
Liriodendron * as shown on pages 404, 405. 
If we added to these examples other magnoliaceous genera we 
should of course introduce some new variations of structure, but 
these would afford us no better family characters. A formula 
typical of the family would still be the same as that given below the 
three genera mentioned. 
Comparing our magnoliaceous formulas with the ranunculaceous 
ones we find that while prevailing features differ—so much so 
indeed as to make it desirable to group the plants in separate fam- 
ilies—the departures from the type in one family often match those 
of the other. — 
There is, however, a general difference, not shown in the 
figures, which serves to separate the two groups. All mem- 
bers of the magnolia family have in the leaf-pulp, floral leaves, 
pith, and other soft parts, minute reservoirs of volatile oil, 
which are entirely lacking in the crowfoot family. These little 
reservoirs may be seen readily with a hand lens by viewing 
1 Ex-stip’u-late < L. ex, without; stipula, stipule. 
2 Hab’it < L. habitus, appearance. 
3The plant formulas referred to in this and succeeding sections, 
together with the ranunculaceous formulas already given, are grouped 
on pages 404-427 to facilitate their being compared with one another. 
