THE ACORNS AND THEIR GERMINATION. 409 
The proportion of the caulicle to the stalks appears to be constant in the same species, as I 
have satisfied myself by examination of numerous acorns of the same species from widely separated 
localities. 
In all the black-oaks which I could study the caulicle is longer than the stalks. [191 (2) ] 
Thus in Quercus nigra, imbricaria, pumila, and Kelloggit I find it twice as long; in Q. 
coccinea and tinctoria, rubra, ilicifolia, and agrifolia, it is three to four times as long. @Q. densiflora, 
of the section Androgyne, is similar in this respect to the first. 
few white-oaks resemble the black-oaks in the proportion of these parts. These are Q. 
macrocarpa and undulata, and especially Q. Robur of Europe. In this and in the Californian Q. chry- 
solepis I tind the caulicle nea:ty three times longer than the stalks; in both of them I also notice 
the plumule unusually developed. 
But in the majority of white-oaks the caulicle is shorter than the stalks of the cotyledons; 
I have seen it in the American Q. alba, stellata, Garryana, Douglasii, Breweri, Prinus, Mihlenbergii, 
prinoides, Michauxit, bicolor, dumosa, pungens} and in Q. occidentalis of southwestern Europe. It 
is very interesting to find that in the hybrids of Q. macrocarpa and alba, which externally resemble 
more the former than the latter, the proportion of the stalks and the caulicle is entirely that of the 
latter and not of the former. I have observed this fact in hybrids from Ilinois (Ha//) as well as 
from Vermont (Pringle). 
By far the longest stalks or — however, are found in @. virens; in this species not only 
the cotyledons, as is well known, but also their stalks, are coalescent: the caulicle itself is very 
short, only about one-fourth or one-fifth the length of the stalks, and the place where they separate 
from the caulicle is indicated by the very small and imperfect plumule, completely imbedded within 
the connate base of the stalks. 
The acorns of all oaks germinate in or on the ground, the thickened stalks and the caulicle elon- 
gate; the former become 2 to 4 or nearly as much as 6 lines long, while the cotyledons themselves 
remain enclosed in the cracked seedshell, and from between the bases of the stalks the 
plumule grows up into the ascending axis, nourished by the food contained in the cotyle- [192 (3) ] 
dons ; these become exhausted and rot away about the end of the first season, while the 
radicle about the same time swells up, evidently absorbing part of the matter contained in them, 
and thus laying up a store of food for the next season. 
The process in Q. virens is essentially the same; it differs somewhat in that the connate stalk 
of the cotyledons remains more slender, but elongates more, mostly to the extent of one inch or even 
more ; the caulicle and upper part of the root swells up at once, while the developing plumule forces 
its way up through a slit in the base of the stalk. It seems that the danger of losing connection 
with the storehouse of the cotyledonous mass through the long and slender passage of the stalk, 
necessitates the transfer of the food-matter to a nearer and safer place of deposit. But why, it may be 
asked, is the connection so much longer and more slender than in other oaks? At all events it suf- 
fices, as long as it is fresh and unimpaired, to carry over in a very short time the starchy and sweet 
contents from the cotyledons to the tuber; and before the ascending axis is an inch high and bears as 
yet only a few minute bracts, the tuber is already forming and it soon reaches the size of the cotyle- 
dons themselves ; it is, however, longer and more slender, of a fusiform shape, about three to four 
lines thick and one to two inches long, attenuated below into the long tap-root. 
1 All Colorado forms of Q. undulata which I have exam- stalks longer than the caulicle. Finding the proportions of 
ined, those with large and deciduous leaves, vars. Gambelit, these parts to be 
form from the edge of the cafion of the Arkansas, which I Q. grisea, as a distinct species, provided other characters can 
stalks, while the pale and usually persistent-leaved forms 
from Arizona, the true Q. pungens of Liebmann, have the 
be found to confirm this view. In a few acorns of this form 
I have seen the cotyledons adhering together, but in the 
majority they were free 
