June 25, 1920] 



SCIENCE 



623 



the manifold diversities of the environment, 

 that is, in the wide view they are adaptive. 

 Natural selection may (or may not) act upon 

 the products of this differentiation as they 

 are formed. Inheritance of acquired char- 

 acters in the ordinary sense of this expression 

 is not implied here, though some recasting of 

 current ideas of the individuality of the germ 

 plasm and the nature of the mechanism of 

 heredity is a necessary consequence of the 

 recent studies in general protoplasmic phys- 

 iology to which reference has been made. 



ISTow this process of senescence of tissue is 

 to a large extent reversible; that is, a special- 

 ized tissue may dedifferentiate and return to 

 the embryonic type, as happens in the ordi- 

 nary processes of reproduction, regeneration 

 and the like. But this capacity for dediffer- 

 entiation is not universal, nor where it occurs 

 is it always accomplished with equal facility. 

 In general, specialized tissues return to the 

 imdifferentiated condition with greater diffi- 

 culty than do the simpler and more general- 

 ized kinds and the capacity for form regula- 

 tion diminishes pari passu with the increase 

 in complexity of bodily organization. In 

 higher organisms groups of general body cells 

 are incapable of reproducing the whole body 

 as in lower forms; in a salamander an entire 

 limb can be regenerated, but in a man this is 

 impossible; and differentiated nerve cells are 

 incapable even of cell division. To this ex- 

 tent, tissue differentiation is irreversible. 



This progressive stabilization of heritable 

 patterns of organization is an essential factor 

 in evolution, and to the extent that these 

 patterns are irreversible the future course of 

 evolution is predetermined. For, given a 

 particular inherited structural pattern, varia- 

 tions will be distributed around this as a 

 mode is accordance with a different frequency 

 curve than would be shown if the inherited 

 pattern were different; and the same applies 

 to mutations..'^ 



An aquatic species which has acquired 

 adaptation to life on land has established new 



7 Metcalf , M. M., ' ' Adaptation through Natural 

 Selection and Orthogenesis," Am. Nat., Vol. 47, 

 1913, pp. 65-71. 



modes around which its variations and muta- 

 tions are distributed. True, it may in time 

 return to the water, though never in a higher 

 animal by the process of dedifferentiation to 

 the original aquatic form but only along lines 

 of further differentiation derived from and 

 congruous with its established terrestrial 

 patterns. 



Again, with the establishment of the ladder 

 type of central nervous system as seen in 

 annelid worms, a stable pattern was laid down 

 with certain functional capacities. On the 

 other hand, with the establishment of the 

 tubular type of central nervous systems in 

 early vertebrate ancestors, a different pattern 

 was fabricated with its own characteristic 

 correlated behavior. On the basis of each of 

 these matured and stabilized tissue differen- 

 tiations a wide variety of central nervous 

 systems has been derived — from the annelid 

 type the whole series of arthropods and from 

 the protochoi'date type the whole series of 

 vertebrates. But throughout each of these 

 phyla the fundamental pattern has not been 

 changed, nor have we any adequate evidence 

 that one has ever been transformed into the 

 other. 



In other words, from the time when these 

 two structural patterns were first established 

 and stabilized, the process was irreversible; 

 the tissues concerned have in this respect and 

 to this extent passed from the " young " or 

 labile state to the " mature " or rigidly deter- 

 mined form. The undifferentiated type ante- 

 cedent to both of these phyla was labile in the 

 sense that under appropriate conditions it 

 could differentiate in either direction; but 

 having passed over to either one of the differ- 

 entiated forms, it has apparently lost its 

 capacity to transform to the other type, either 

 by dedifferentiation and remodelling of its 

 pattern or by any other method. At any rate 

 we have no convincing evidence that this has 

 been done. In short, the whole future course 

 of evolution of the vertebrate phylum was set 

 in a different direction from that of the 

 arthropods from the first appearance of a 

 neural tube. 



