370 ON TWO NEW CRYSTALLINE COMPOUNDS 



have that definiteness of character which is claimed for those of the chemical law. 

 The planetary orbits are not perfect ellipses. The ratios of the harmonic scale are but 

 approximatively realized. The arrangement of leaves on the stem is not perfectly regu- 

 lar. Isomorphism is seldom absolute. In all we observe only a tendency towards a 

 maximum effect, which is the perfect expression of the law, but which is rarely fully 

 reached. The limits of variation are broader in some instances than in others, but Ave 

 find no case in which there is absolutely none. This same character, which pervades 

 the other phenomenal laws of nature, I claim for the great law of Chemistry. The 

 definite proportion I regard as a maximum towards which the chemical force strives, 

 a maximum from which the deviations in most cases are small, although in others 

 they may be very large ; and I maintain that this view of the subject, which the 

 memoir has aimed to establish, is supported by the analogies of nature. 



When the dynamical law has been discovered, of which the phenomenal law was 

 merely the outward manifestation, as Kepler's laws were merely the phenomena of the 

 law of universal gravitation, the very variations have been seen to be necessary con- 

 sequences of the law itself; and if ever the dynamical law which governs chemical 

 phenomena shall be discovered, it is most probable that the variations from the law of 

 definite proportions will become as much a matter of calculation as the perturbations 

 of astronomy. In both cases the perturbation is apparently due to the influence of an 

 extraneous mass of matter. 



The argument from analogy becomes stronger when Ave consider the equivalent num- 

 bers. I have shown in a former memoir,* that these numbers may be connected by a 

 very simple numerical laAV ; but here, as in other cases, we find merely a tendency to- 

 wards the law, not an absolute agreement with it, the difierences between the theoreti- 

 cal and the experimental equivalents being in many cases too great to be covered by 

 errors of observation. The present memoir may throw light upon these discrepancies ; 

 for, to say the least, it is possible that the differences may originate in variations of the 

 equivalent itself, and that the theoretical equiA^alent may be the maximum towards 

 which the chemical force tends. On comparing carefully the different determinations 

 of the chemical equivalents, many facts will be noticed supporting this view ; those 

 equivalents, for examj)le, which coincide Avith, or very nearly approach, whole numbers, 

 such as those of oxygen, carbon, and sulphur, will be found as a general rule to have 

 been determined by the analysis or synthesis of compounds Avhose elements are united by 

 a strong chemical force ; also, when the equivalents have been determined by essentially 



* This volume, page 235. 



