Contractions, however, certainly may be excited in different parts of a frog, without making any division of its skin, by laying the part of the frog to be excited upon a plate of zinc, or tin-foil, and passing a piece of silver over it, till all three are in contact with each other[7]. Yet even here the influence does not pass into the part in so diffused a state as it may at first appear to do. For the skin of these animals is abundantly supplied with nerves, whose trunks communicate, at different places, with those which supply the muscles. And the contractions are always strongest, and most readily excited, when the silver is passed over the course of any of the nerves, which go to the muscles. From the fact, which I have before mentioned, that a limb may be made to contract, when the metals have apparently no communication with any part of it except its nerve; it might reasonably be doubted, whether, in any case, a communication between the muscles, as well as the nerve, and the metals, were necessary, in order that contractions may be excited. Several considerations, however, induce me to believe, that such communication is absolutely requisite. If the contact of two different metals were alone sufficient to excite contractions; contractions should always take place, whenever a good conductor is interposed between the metals, and the nerve alone. But I have, in no instance, observed this to be the case. In the experiment, where the crural nerve must be supported upon a silver probe, it is necessary that the piece of silver, with which the zinc is brought in contact, should communicate either immediately, or through some good conducting medium, with the muscles of the foot, or leg, before any contraction takes place. And even in the experiment, where water forms the only communication between the metals, and the origin of the sciatic nerves, that same water, it must be observed, forms likewise a communication between the metals and the muscles, to which these nerves are distributed. But the fact, which appears to me most decisive of this question, is the following: When a nerve, which for some time has been detached from surrounding parts, is either carefully wiped quite dry with a piece of fine muslin, or (lest this should be thought to injure its structure,) suffered to remain suspended till its moisture has evaporated; no contractions can be excited in the muscles, to which it is distributed, by touching it alone with any two metals in contact with each other. But, if it be again moistened with a few drops of water, contractions instantly take place: and, in this way, by alternately drying and moistening the nerve, contractions may, at pleasure, be alternately suspended and renewed for a considerable time. It may, indeed, be contended, that the moisture softened, and thus restored electricity and free expansion to the dried cellular membrane surrounding the fibres, of which the trunk of a nerve is composed; and