§ 102. Another condition to Evolution, of the same order as the last though of a different genus, must be noted. We have found that permanent re-arrangement among the units of an aggregate, can take place only when they have neither extreme immobility nor extreme mobility. The mobility and immobility thus far considered (at least in all aggregates except social ones) are those due to mechanical cohesion. There is, however, what we must call chemical cohesion, which also influences the mobility of the units, and consequently the re-arrangement of them. Manifestly, if two or more kinds of units contained in any aggregate, are united by powerful affinities, an incident force, failing to destroy their cohesions, will not cause such various re-arrangements as it would, could it produce new chemical combinations as well as new mechanical adjustments. On the other hand, chemical affinities that are easily overcome, must be favourable to multiplied re-arrangements of the units. This condition, as well as the preceding one, is fulfilled in the highest degree, by those aggregates which most variously display the transformation of the uniform into the multiform. Organic bodies are on the average distinguished from inorganic bodies, by the readiness with which the compounds they consist of undergo decomposition, and recomposition: the chemical cohesions of their components are so comparatively small, that small incident forces suffice to overcome them and cause transpositions of the components. Further, between the two great divisions of organisms, we find a contrast in the degree of Evolution co-existing with a contrast in the degree of chemical modifiability. As a class, the nitrogenous compounds are peculiarly unstable; and, speaking generally, these are present in much larger quantities in animal tissues than they are in vegetal tissues; while, speaking generally, animals are much more heterogeneous than plants. Under this head it may be well also to point out that, other things equal, the structural variety which is possible in any aggregate, must bear a relation to the number of kinds of units contained in the aggregate. A body made up of units of one order, cannot admit of so many different re-arrangements, as one made up of units of two orders. And each additional order of units must increase, in a geometrical proportion, the number of re-arrangements that may be made. * * * * *