1318. Thus the whole seed, shell and embryo, completely resembles a pinnate leaf with a phyllodium, as we see it in the umbellate plants, only it must be thought of as one so involuted, that the fine leaves adhere reversed in the phyllodium. Seeds may consequently change into leaves. A seed has therefore been formed also in all its parts like a papilionaceous corolla. This resemblance speaks moreover retrogressively, for the petals of the papilionaceous corolla being viewed only as a single leaf-bud. Seeds may therefore change also into corollæ. All parts of the seed are thus an unity, a single pinnate leaf, and it is consequently impossible for them to have been patched up out of what has been called the seed-ovum, namely, the testa and embryo, which would come from some other source or out of the pollen. 1319. The seed is the whole plant in miniature: the root being portrayed in the umbilical cord, and radical leaf in the phyllodium; stalk in the radicle; caudal leaves in the seed-lobes; ramuscule in the cotyledonal petiole; ramular leaves in the cotyledons. Seeds may thus change into an entire plant. The seed is consequently nothing new in the plant, but the repetition of the same under the relations and forms of the root. 1320. It is plain that the seeds must always change into the same plant; for they are indeed nothing else but this. The identity ensuing upon propagation is accordingly nothing singular and incomprehensible; it would be so were it otherwise. With the seed the plant has but reverted again to its primary condition, to the galvanic, mucous vesicle, out of which in a secondary manner the young plant is developed like as is the first plant out of the primary vesicle. 1321. The radicle is not therefore root itself, but only emits rootlets. 1322. The germ or the radicle must observe different positions towards the umbilicus, according as the seed-leaf or the testa has been more or less involuted, and according as the germ is removed from the micropyle. 1323. The albumen or perisperm is no particular organ, but only the deposit from the sap, which the inner wall of the testa secretes. The albumen stands in no organic connexion with the parts of the seed. That therefore which has become connate with the kernel or nucleus cannot be albumen. 1324. The arillus can be none other than bud-seeds of the testa, because it is placed under the phyllodium. It corresponds to the floral involucrum or bracteal scales. 1325. As the seeds are none other than leaves that have remained stationary in the condition of root, so must they pass through the three leaf-stages. There can be therefore only three principal differences in the seed-formation. 1326. The seeds of plants with reticular leaves consist of several leaves arranged symmetrically or in pairs. They have necessarily _two_ seed-lobes--Dicotyledones.