420 



SCIENCE. 



[N. S. Vol. XX. No. 509. 



Avogadro and Ampere, and under the 

 similarly fruitful laws of gases established 

 by Dalton and Gay-Lussae. Perhaps the 

 most striking feature of this progress, in a 

 general way, is the gradual disappearance 

 it has entailed of the imaginary lines which 

 have been long thought to separate the 

 fields of chemistry and physics. Through 

 the remarkable discoveries of Faraday the 

 two fields have been found to overlap in 

 actual electrical contact. Through the 

 wonderful revelations of spectrum analysis, 

 originating with Bunsen and Kirchhoff, 

 they have been proved to be very largely 

 common ground. And through the broader 

 generalizations inaugurated by Willard 

 Gibbs, Helmholtz and others, they are now 

 both somewhat in danger of being annexed 

 as a subprovince of rational mechanics. 



To one whose work has fallen more espe-, 

 cially in the fields of precise astronomy, 

 geodesy or metrology, it might seem a 

 just reproach to chemistry that it is a sci- 

 ence whose measurements and calculations 

 demand, as a rule, no greater arithmetical 

 resources than those of four-place tables 

 of logarithms and anti-logarithms. The 

 so-called ' Constants of Nature ' supplied by 

 chemistry are, in fact, known with a low 

 degree of certainty, a degree expressed, 

 say, by three to five significant figures. A 

 small amount of reflection, however, will 

 convince one that the phenomena with 

 which the chemist has to deal are usually 

 far more complex than those which have 

 yielded the splendid precision of astron- 

 omy, geodesy and metrology. Moreover, it 

 should be observed that the certainties even 

 of these highly perfected sciences are very 

 unequal in their different branches. It 

 appears more correct, therefore, as well as 

 more just, considering the central position 

 it occupies and the wide range of its rami- 

 fications, along with the vast aggregate of 

 qualitative and quantitative knowledge it 

 has massed, to assert that the precision of 



chemistry affords the best numerical index 

 of the present state of physical science. 

 That is, when reduced to the most compact 

 form of statement, the -certainties of phys- 

 ical science are best indicated, in a general 

 way, by a table of the combining weights 

 of the eighty-odd chemical elements. 



When one contemplates the numbers of 

 such a table, and when one adds to its sug- 

 gestions those which flow from the various 

 periodic groupings of the same numbers, 

 he can hardly avoid being inspired by the 

 day dreams of those who have looked long 

 for the atomic unity of matter. But how- 

 ever the grand problem which thus ob- 

 trudes itself may be resolved finally, it ap- 

 pears certain that this table must stand as 

 one of the great landmarks along the path 

 of progress in physical science. 



It was justly remarked by Laplace in his 

 'Systeme du Monde' that 'L 'Astronomic, 

 par la dignite de son objet et par la per- 

 fection de ses theories, est le plus beau 

 monument de I'esprit humain, le titre le 

 plus noble de son intelligence ' ; and we 

 must all admit that subsequent progress 

 has gone far to maintain this high position 

 for the most ancient and interesting of the 

 older sciences. One finds little difficulty in 

 accounting for the early rise of astronom- 

 ical science and for the universal interest 

 in celestial phenomena. Their immanence 

 and omnipresence appeal even to the dullest 

 intellects. But it is not so easy to account 

 for the remarkable fact that although as- 

 tronomy deals chiefly with the relations of 

 bodies separated by immense distances, 

 progress in its development has t^us far 

 been at least equal to, if not in advance of, 

 the progress of physics and chemistry, 

 which have to deal with matter close at 

 hand. Without attempting a full explana- 

 tion of this fact, it may suffice to observe 

 that the principal phenomena of astronomy 

 thus far developed appear to be relatively 

 simple in comparison with those of the 



