In _R. raphanistrum_ the siliqua is articulated, that is to say, contracted at intervals, and the seeds placed each in a division. In _R. sativus_ the siliqua is continuous, and forms a single cavity. Some botanists had made this difference the basis of two distinct genera, _Raphanistrum_ and _Raphanus_. But three accurate observers, Webb, Gay, and Spach, have noticed among plants of _Raphanus sativus_, raised from the same seed, both unilocular and articulated pods, some of them bilocular, others plurilocular. Webb[42] arrived at the same results when he afterwards repeated these experiments, and he observed yet another fact of some importance: the radish which sows itself by chance, and is not cultivated, produced the siliquæ of _Raphanistrum_.[43] Another difference between the two plants is in the root, fleshy in _R. sativus_, slender in _R. raphanistrum_; but this changes with cultivation, as appears from the experiments of Carrière, the head gardener of the nurseries of the Natural History Museum in Paris.[44] It occurred to him to sow the seeds of the slender-rooted _Raphanistrum_ in both stiff and light soil, and in the fourth generation he obtained fleshy radishes, of varied colour and form like those of our gardens. He even gives the figures, which are really curious and conclusive. The pungent taste of the radish was not wanting. To obtain these changes, Carrière sowed in September, so as to make the plant almost biennial instead of annual. The thickening of the root was the natural result, since many biennial plants have fleshy roots. The inverse experiment remains to be tried—to sow cultivated radishes in a poor soil. Probably the roots would become poorer and poorer, while the siliquæ would become more and more articulated. From all the experiments I have mentioned, _Raphanus sativus_ might well be a variety of _R. raphanistrum_, an unstable variety determined by the existence of several generations in a fertile soil. We cannot suppose that ancient uncivilized peoples made essays like those of Carrière, but they may have noticed plants of _Raphanistrum_ grown in richly manured soil, with more or less fleshy roots; and this soon suggested the idea of cultivating them. I have, however, one objection to make, founded on geographical botany. _Raphanus raphanistrum_ is a European plant which does not exist in Asia.[45] It cannot, therefore, be this species that has furnished the inhabitants of India, China, and Japan with the radishes which they have cultivated for centuries. On the other hand, how could _R. raphanistrum_, which is supposed to have been modified in Europe, have been transmitted in ancient times across the whole of Asia? The transport of cultivated plants has commonly proceeded from Asia into Europe. Chang-Kien certainly brought vegetables from Bactriana into China in the second century B.C., but the radish is not named among the number. +Horse-radish+—_Cochlearia Armoracia_, Linnæus.