The animal kingdom presents two relatively sudden and very striking advances in complexity and in the size of the unit upon which natural selection acts unmodified. These advances consist in the aggregation of units which were previously independent and exposed to the full normal action of natural selection, and the two instances are, of course, the passage from the unicellular to the multicellular, and from the solitary to the social. It is obvious that in the multicellular organism individual cells lose some of the capacities of the unicellular—reproductive capacity is regulated and limited, nutrition is no longer possible in the old simple way and response to stimuli comes only in certain channels. In return for these sacrifices we may say, metaphorically, that the action of natural selection is withdrawn from within the commune. Unfitness of a given cell or group of cells can be eliminated only through its effect upon the whole organism. The latter is less sensitive to the vagaries of a single cell than is the organism of which the single cell is the whole. It would seem, therefore, that there is now allowed a greater range of variability for the individual cells, and perhaps, therefore, an increased richness of the material to be selected from. Variations, moreover, which were not immediately favourable would now have a chance of surviving. Looked at in this way, multicellularity presents itself as an escape from the rigour of natural selection, which for the unicellular organism had narrowed {19} competition to so desperate a struggle that any variation outside the straitest limits was fatal, for even though it might be favourable in one respect, it would, in so small a kingdom, involve a loss in another. The only way, therefore, for further advantageous elaboration to occur was by the enlargement of the competing unit. Various species of multicellular organisms might in time be supposed in turn to reach the limit of their powers. Competition would be at its maximum, smaller and smaller variations would be capable of producing serious results. In the species where these conditions prevail an enlargement of the unit is imminent if progress is to occur. It is no longer possible by increases of physical complexity and the apparently inevitable sequence is the appearance of gregariousness. The necessity and inevitableness of the change are shown by its scattered development in very widely separated regions (for example, in insects and in mammals) just as, we may suspect, multicellularity appeared.