According to the conception I have been explaining, cells merge their independent individuality in that of the whole, and so the force that directs their ultimate development, and that leads to their appropriate elaboration, cannot be within them, cannot reside in special groups of determinants, in the sense of Weismann. It is given by the relations in which the cells come to stand to the whole organism and to the various parts of the organism, and, on the other hand, to surrounding things. Naturally, such relations differ with the place or position occupied by cells in the whole organism, and in this way there come to be innumerable conditions making for diverging directions of development, for division of labour, and for dissimilar, histological differentiation. The part played by a cell, as Vöchting puts it, will depend upon the position it comes to assume in the whole living unit. To use an expression of Driesch's, dissimilar differentiation of cells is a 'function of position.' Such a conception my brother and I, in our Studies on the Germ-layer Theory, sought to establish clearly by many examples from the histology of the coelenterates and of higher animals; such a conception for long has been clearly expressed in physiological botany.
The simpler nature of plants in structure and function makes it easy to conduct experimental observations upon this point.
I have already described how either side of the prothallus of a fern may be made to produce male or female organs, according as it is kept in the light or in the dark. Similarly, taking a willow slip, roots may be made to appear at one end by moisture and darkness, while they will not appear on the end kept in the light.
The experiments of botanists and of fruit-growers show that young buds and the rudiments of roots are indifferent structures, the further growth of which depends entirely upon the conditions in which they are placed. 'One and the same bud may grow to a long or short vegetative shoot, to a floral shoot, to a thorn, or may remain undeveloped. The same root rudiment may grow to a main tap-root or may form a secondary lateral root. The conditions that determine the mode in which these structures will develop are quite within the power of the experimenter. We have shown already and could show further, that he is able to determine the mode of growth by cutting, bending, tying in a horizontal position, and so forth: For such reasons, Vöchting describes plants as masses of tissue, practically plastic, and which may be moulded at the discretion of the investigator. 'For instance, in the case of Prunus spinosa, a branch may be produced in place of a thorn by cutting a growing shoot at the proper height, in spring. The buds below the point where the cut was made turn to shoots like the rest of the plant and complete the interrupted growth, while on an uncut stem they would have grown to thorns. Thus, the rudiment of a thorn has been changed to that of a shoot' (Vöchting).
Although it is more difficult to carry out experiments upon animals, some good instances are known. If a piece cut from the stem of Antennularia (a hydroid polyp) be placed vertically, in a short time new branches and new 'roots' spring from it. In this case, again, the position of the new growths is determined by the relation in which the stem is placed to gravity. 'The tentacles arise only at the end turned towards the zenith; the "roots" from the parts directed towards the ground' (Loeb).
A similar example may be taken from among vertebrates. The notochord arises from a set of cells which are in close relation with the fused tips of the blastopore. By exposing developing frog's eggs to abnormal conditions, I was able, in some cases, to produce a hypertrophy of one of the lips of the blastopore. When fusion of the lips took place the normal lip united with the rim of the protruding hypertrophied lip. As a result of this the notochord and the nerve plate came to arise, not from the usual set of cells, but from those cells that, by the abnormal condition, had come to lie in the place for the notochord. The protruding cells, which normally would have developed into notochord and nerve plate, grew into a simple fold of the external skin.
Moreover, it is well known in pathology that mucous membranes may lose their proper character and assume the qualities and aspect of the external skin, when, as in cases of prolapse, fistula, etc., they have been exposed for some time to the air.
The relations of different parts to each other and to the whole are known as correlations. Correlation exists in all the stages of the development of an organism, sometimes in one way, sometimes in another. One must note very carefully that Weismann's doctrine of determinants, according to which all that happens in development follows a prearranged plan, is entirely in opposition to this correlative character of the changes that occur during development.
Here I shall give a few quotations from botanical and zoological writers:
'If the stem of a plant be cut so that it retains its roots, but is deprived of leaves and shoots, then the adventitious buds will produce new leaves and shoots. If, however, the stem be cut so as to deprive it of roots, then the same cells that in the other case produced leaves and shoots will now produce roots. Precisely the same occurs with a piece of the root. In fact, it appears as if the idioplasm knew what parts of the plant were wanting, and what it must do to restore the integrity and vital capacity of the individual.' 'The idioplasm in the remaining part of a plant must be affected when an important part has been removed, because the idioplasm of the lost part is no longer capable of having influence.' 'It is clear enough that necessity acts as a stimulus, and that each definite need calls into existence the appropriate reaction.'