222 THE. FORMATION 
Liliaceae, Primulaceae, Rubiaceae, etc.; (3) protective cells present in the 
seed coats, endosperm little or none, Boraginaceae, Crassulaceae, Cruciferae, 
Labiatae,, Papilionaceae, etc.; (4) protective elements present, Ascleptas, 
Campanula, Gentiana, Silene, Saxifraga, etc.; (5) protective cells present, 
endosperm thick-walled, Euonymus, Helianthemum, Ribes.. The protective 
cells are of various kinds: (1) epidermal cells strongly cuticularized, 
Caryophyllaceae, Crassulaceae, Fumariaceae, Saxifragaceae; (2) paren- 
chyma thick-walled, several-layered, Aesculus, Castanea, Fagus; (3) 
parenchyma cells with the inner or radial walls thickened, Campanula, 
Erythraea, Gentiana; (4) epidermal cells cup-shaped, thick-walled, Cruci- 
ferae, Ribes, Vaccinium; (5) parenchyma with thickened, cellulose walls, 
Geranium, Viburnum, (6) a single row of stone-cells, Labiatae; (7) tissue 
of stone-cells, Hippuris, Naias, Potamogeton; (8) elongate stone-cells, 
Coniferae, Cupuliferae, Euphorbia, Linum, Malva, Viola; (9) short, colum- 
nar, thick-walled branched cells, Cucurbitaceae, Datura, Hypericum; (10) 
prosenchyma with cellulose walls, Clematis; (11) prosenchyma with lignified 
walls, Fravinus, Rhamnus, Ranunculus. The seed-coats have a certain in- 
fluence in determining germination at the proper time, inasmuch as they 
make it difficult for the seed to germinate under the stimulus of a quantity 
of warmth and moisture insufficient to support the seedling. The effect of 
the endosperm, as well as that of other food-supply in the seed, upon’ 
germination and the establishment of the seedling is obvious. 
The behavior of seed or spore with respect to germination depends in 
a large degree upon the character of the protoplasm or embryo, though in 
just what way is at present a matter of conjecture. It is evident that many 
seeds are not viable because fertilization has not been effected, and in con- 
sequence no embryo has developed. This is the usual explanation of the 
low germinating power of the seeds of some species, especially polysperma- 
tous ones. But even in viable seeds the behavior is always more or less 
irregular. The seeds of some species will grow immediately after ripening, 
while others germinate only after a resting period of uncertain duration. 
The same is true of spores. Even in the case of seeds from the same 
parent, under apparently similar conditions, while the majority will germi- 
nate the first year, some will lie dormant for one or more years. The 
precise reason why many seeds and spores germinate more readily after 
being frozen is equally obscure. The period of time for which disseminules 
may remain viable is extremely diverse, though, as would be expected, it 
is much longer as a rule for seeds than for spores. The greater vitality of 
seeds in the case of ruderal plants suggests that this diversity may be due 
simply to variation in the vigor of the embryos. It would seem that under 
proper conditions seeds may retain their viability for an indefinite period. 
