Coming to the observation of what happens under natural conditions, we are struck at the outset by the fact that the inheritance of acquired characteristics, if it works at all, must work under some system of salutary control and not as a blind physiological law. For if each generation starts with some measure at least of what the former generation had acquired, and adds to it by its own activity, then all acquired characteristics would ere long attain a monstrous development, and the species would perish under them. But nothing of the kind is observed to happen. The continual use of the muscles in the labouring classes has not made men stronger than they were thousands of generations ago. The habit of handling the spade and hoe has never produced a peasant child born with callosities on its hands. The horn of the rhinoceros, which on Lamarckian principles we must regard as developed by the gradual increase of a callosity formed by grubbing for roots, does not grow beyond a certain size, however the species may go on grubbing. The Lamarckian law, then, if it has any real effect at all, can only express half the truth about the action of heredity on acquired characteristics. As the column of water in a fountain hovers about a certain height, so the action of heredity in the accumulation of the effects produced by the use of organs seems to have a limit beyond which it cannot pass. May it not be that heredity is really as false an expression for the phenomenon as the popular superstition about ‘water seeking its own level’ is for the upspringing of a fountain?
The cases of co-adaptation, where one organ appears to be developed by use and others by the use of that, as in the case of the Irish elk referred to above, are met by instances just as striking where the elements of modification by use cannot come into play. Weismann mentions the case of the ingenious brush arrangement on the anterior legs of the bee, which the insect uses for cleansing its antennæ. Two adaptations are here developed—a little semicircular notch in the leg, set with small bristles, and a movable projection or flap used for pressing the antenna into the notch as it is drawn through. The bee, no doubt, would naturally try to clean its antennæ with its fore-legs, but how could this process develop the special arrangements referred to in the hard or scaly covering of its limbs? It is not until the shell of the insect has grown quite hard and incapable of further vital changes that the arrangement comes into use. Again, the stridulating noise produced by the legs of the grasshopper is due to serrations occurring on different joints of the limb. Serrations on one joint would in no way tend to develop them on the other, but rather the contrary, yet there they are, in harmonious co-operation. If Nature can obtain these effects, as she does in numberless instances, without the aid of Lamarck’s principle, we cannot help asking whether that principle is ever operative at all.