primitive a mechanism as this could be easily acquired by any soft-bodied animal like a snail. Besides, we have many intermediate links between the pond-snails and my cyclostoma here. There are some species which live in moist moss, or the beds of trickling streams. There are others which go further from the water, and spend their days in damp grass. And there are yet others which have taken to a wholly terrestrial existence in woods or meadows and under heaps of stones. All of them agree with the pond-snails in having an operculum, and so differ from the ordinary land and river snails, the mouths of whose shells are quite unprotected. Thus land-nails have two separate origins--one large group (including the garden-snail) being derived from the common fresh-water mollusks, while another much smaller group (including the cyclostoma) is derived from the operculated pond-snails. How is it, then, that naturalists had so long overlooked this distinction? Simply because their artificial classification is based entirely upon the nature of the breathing apparatus. But, as Mr. Wallace has well pointed out, obvious and important functional differences are of far less value in tracing relationship than insignificant and unimportant structural details. Any water-snail may have to take to a terrestrial life if the ponds in which it lives are liable to dry up during warm weather. Those individuals alone will then survive which display a tendency to oxygenise their blood by some rudimentary form of lung. Hence the possession of lungs is not the mark of a real genealogical class, but a mere necessary result of a terrestrial existence. On the other hand, the possession of an operculum, unimportant as it may be to the life of the animal, is a good test of relationship by descent. All snails which take to living on land, whatever their original form, will acquire lungs: but an operculated snail will retain its operculum, and so bear witness to its ancestry; while a snail which is not operculated will of course show no tendency to develop such a structure, and so will equally give a true testimony as to its origin. In short, the less functionally useful any organ is, the higher is its value as a gauge of its owner's pedigree, like a Bourbon nose or an Austrian lip. XIX. _DOGS AND MASTERS._