physiological units [or constitutional units as they might otherwise be called]--each kind of organism being built up of units peculiar to itself. We recognized in these units, powers of arranging themselves into the forms of the organisms to which they belong; analogous to the powers which the molecules of inorganic substances have of aggregating into specific crystalline forms. We have consequently to regard this proclivity of the physiological units, as producing, during the development of any organism, a combination of internal forces that expend themselves in working out a structure in equilibrium with the forces to which ancestral organisms were exposed; but not in equilibrium with the forces to which the existing organism is exposed, if the environment has been changed. Hence the problem in all cases is, to ascertain the resultant of internal organizing forces, tending to reproduce the ancestral form, and external modifying forces, tending to cause deviations from that form. Moreover, we have to take into account, not only the characters of immediately-preceding ancestors, but also those of their ancestors, and ancestors of all degrees of remoteness. Setting out with rudimentary types, we have to consider how, in each successive stage of evolution, the structures acquired during previous stages have been obscured by further integrations and further differentiations; or, conversely, how the lineaments of primitive organisms have all along continued to manifest themselves under the superposed modifications. §179. Two ways of carrying on the inquiry suggest themselves. We may go through the several great groups of organisms, with the view of reaching, by comparison of parts, certain general truths respecting the homologies, the forms, and the relations of their parts; and then, having dealt with the phenomena inductively, may retrace our steps with the view of deductively interpreting the general truths reached. Or, instead of thus separating the two investigations, we may carry them on hand in hand--first establishing each general truth empirically, and then proceeding to the rationale of it. This last method will, I think, conduce to both brevity and clearness. Let us now thus deal with the first class of morphological problems. [NOTE.--In preparation for treating of morphological development, sundry other general considerations should have been included in the foregoing chapter when originally published. This seems the most appropriate place for now naming them. Some were implicitly contained in the first volume, but it will be well definitely to state these, as well as the others not yet implied.