COLLIGATIVE RELATIONS AND SCIENTIFIC LAWS 131 



tion; or it may first appear as a theoretic deduction, as did in part 

 Moseley's law, and only subsequently be shown to describe and pre- 

 dict experience. Whichever link is formed first, we rest dissatisfied 

 until both have been forged. Without the link to experience there is 

 no colligative relation. Without the link to theory it remains "unra- 

 tionalized" and, so, hard to "grasp" and manage with the manifold 

 others of its kind. 



Even strikingly efficient colligative relations may then leave us 

 dissatisfied as long as they remain uncorrelated by any theory. 

 Evoking a determined effort to provide theoretic accommodation 

 for all "purely empirical" relations, this healthy sense of dissatisfac- 

 tion supplies an indispensable element in the internal dynamics that 

 renders the scientific movement self-preservative. For science would 

 surely and rapidly degenerate into natural history— into a chaotic 

 and unmanageable welter of purely empirical relations— were it not 

 for the characteristically strong and enduring endeavor that, in ma- 

 ture sciences, has kept to an absolute minimum the number of these 

 uncorrelated relations. 



A highly efficient relation, one that cannot possibly be dismissed as 

 something that "just happens," will of course be used, preser\ed, and 

 defended in any case. Thus, having lost its theoretic accommodation, 

 the relation of metals survives until it finds a new one. And thus, 

 though it had never any theoretic accommodation, the relation of 

 inertial and gravitational masses survives during the more than two 

 centuries that separated the work of Newton from that of Einstein. 

 But in marginal cases our opinion of a given relation may be de- 

 cisively influenced, in a way not so far indicated, by the extent to 

 which it has (or has not) achieved theoretic accommodation. With 

 faith in the principle of intelligibility, we believe that for all genuine 

 relations such accommodation must be possible. Failing reasonably 

 promptly to achieve accommodation of some relation of doubtful 

 status, we will be very strongly tempted to dismiss it as an "accident" 

 rather than permit the potentially disastrous accumulation of a mul- 

 titude of unaccommodated relations. 



As is clearest in solar eclipses, the ratio of the diameters of moon 

 and sun as seen from the earth is almost exactly 1:1. This is certainly 

 a relation of observables, and was at one time held significant. But 

 tliis relation, unattractive in its meagre generality, is also one we find 

 tlieoretically inexplicable; and we simply, and probably quite prop- 



