by offering to snails full-grown leaves and young, tender ones of the same grass. The latter will be taken and the others left. But if, by a method of cultivation proposed by Sachs, we make a normally siliciferous plant grow where it can get no silica, it will be devoured at once.
Some plants, that were avoided after treatment with alcohol as well as before it, were found to contain a gum which the alcohol failed to remove, and which stood between the snails and the edible substance. Among these were linden, althea, cactuses, and gummy roots. Another series of plants, including an Arum, narcissus, leucojum, and the balsam touch-me-not, which contain no tannin or gum or substances of disagreeable taste or smell, appeared to be protected by raphides. Tabernæmontanus recognized in 1587 that the leaves of these plants produce a violent sensation of burning in the bronchial tubes, and that it is not due to soluble products or juices, but to the raphides, which are abundant in their tissues. This is proved by the fact that the filtered juice of the pounded leaves does not produce the burning sensation, while the residue on the walls of the filter, and the pounded leaves themselves, produce the characteristic sensation that is felt after chewing the fresh leaves. It is also confirmed by the fact that if the leaves of Arum maculatum, for example, are treated with dilute hydrochloric acid, which dissolves the raphides, animals will readily eat them, while they let alone leaves treated with alcohol, even when they have been steeped in sugar-water. In the case of the squill, snails avoid the outside of the scales, which are rich in raphides, and eat the inner sides, which are free from them. So in the narcissus and orchids, and various other plants, there are parts protected by raphides which are objectionable to snails, and other parts free from them that they eat. But, while raphides protect against some animals, they do not against all. Birds and ruminants do not object to the plants containing them; and even snails manifest different degrees of aversion to them. In a similar manner to these plants with raphides, some species of iris are protected by crystals of oxalic acid. It is very probable that the kinds of armor that we have named are available for protection against other animals than snails. But investigation on this subject has not been sufficiently advanced to permit of definite conclusions or generalizations.
Of the kinds of defense named, a minority of the plants studied by M. Stahl possess but one; many are endowed with two; and some with three—as, for instance, Oxalis (oxalic acid, tannin, and hairs); Circæa (bitter hairs, tannin, and raphides); Smilax (thorns, raphides, and poisons); Aloe (leaf-teeth, raphides, and bitter substance); and Pontederia (crystals of oxalate of lime, raphides, and tannin). In fact, considering the number of ene-
