Now this comparative likeness of the molecules is accompanied by difficultyin segregating them. If the mixed fluid is unduly heated, much water distilsover with the alcohol: it is only within a narrow range of temperature thatmolecules of the one kind are driven off rather than the others; and eventhen not a few of the others accompany them. The most interesting and instructiveexample, however, is furnished by certain phenomena of crystallization. Whenseveral salts that have little analogy of constitution, are dissolved inthe same body of water, they are separated without much trouble, by crystallization: subject as they are to uniform forces, they segregate. The crystals of eachsalt do, indeed, usually contain certain small amounts of the other saltspresent in the solution; but from these they are severally freed by repeatedre-solutions and crystallizations. Mark now, however, that the reverse isthe case when the salts contained in the same body of water are chemicallyhomologous. The nitrates of baryta and lead, or the sulphates of zinc, soda,and magnesia, unite in the same crystals; nor will they crystallize separatelyif these crystals be dissolved afresh, and afresh crystallized. On seekingthe cause of this anomaly, chemists found that such salts were isomorphous-thattheir molecules, though not chemically identical, are identical in the proportionsof acid, base, and water, composing them, and in the crystalline forms theyassume when uniting. Here, then, we see clearly that units of unlike kindsare selected out and separated with a readiness proportionate to the degreeof their unlikeness.
There is a converse cause of segregation which it is needless here totreat of with equal fullness. If different units acted on by the same force,must be differently moved; so, conversely units of the same kind must bedifferently moved by different forces. Supposing some group of units formingpart of a homogeneous aggregate, are unitedly exposed to a force which isunlike in amount or direction to the force acting on the rest of the aggregate,then this group of units will separate from the rest, provided that, of theforce so acting on it, there remains any portion not dissipated in molecularvibrations or absorbed in producing molecular rearrangements. After all thathas been said above, this proposition needs no defence.
Before ending our preliminary exposition, a complementary truth must bespecified; namely that mixed forces are segregated by the reaction of uniformmatters, just as mixed matters are segregated by the action of uniform forces.
Of this truth a complete and sufficient illustration is furnished by thedispersion of refracted light. A beam of light, made up of ethereal undulationsof different orders, is not uniformly deflected by a homogeneous refractingbody; but the different orders of undulations it contains are deflected atdifferent angles: the result being that these different orders of undulationsare separated and integrated, and so produce the colours of the spectrum.
A segregation of another kind occurs when rays of light traverse an obstructingmedium. Those which consist of comparatively short undulations are absorbedbefore those which consist of comparatively long ones; and the red rays,which consist of the longest undulations, alone penetrate when the obstructionis very great. How, conversely, there is produced a separation of like forcesby the reaction of unlike matters, is also made manifest by the phenomenaof refraction; since adjacent and parallel beams of light, falling on, andpassing through, unlike substances, are made to diverge. §164. In vague ways the heavenly bodies exemplify that cause of materialsegregation last assigned -- the action of unlike forces on like units.
I say in vague ways because our Sidereal System displays more of aggregationthan of segregation. That the irregular swarms of stars constituting theMilky Way with its branches and gaps and denser regions, have been gatheredtogether from a more widely diffused state, may be reasonably inferred; thoughas we know nothing of the preceding distribution such a change cannot beproved: still less can there be proved a segregative process.
It is true that in clusters of stars, beginning with those having membersconsiderably dispersed and ending with those having members closely concentrated-globularclusters -- we see strong evidence of aggregation; and it may be contendedthat since the mutual gravitations of the stars forming a cluster, differin their degrees and directions from those of the stars from which they haveseparated, there is a kind of segregation. But it must be admitted that theconformity to the above-named principle is but an indefinite one.
There are, however, two classes of facts which exhibit segregation, thoughthey leave us ignorant of its causes. The first is that star-clusters areabundant along the course of the Milky Way: by far the larger number of themlying in the neighbourhood of its plane and relatively few in regions oneither side. The second is that, contrariwise, the nebulae are sparsely scatteredin and about the galactic circle and are relatively numerous in the spacesremote from it. Though there are thus presented two cases of segregationthere is no evidence that these different classes of bodies have been separatedfrom a mixed assemblage, nor is there any indication of the forces by whichthis contrast in distribution has been produced. We can only say that thefacts are congruous with the belief that segregation, probably indirect ratherthan direct in its cause, has been going on.
