170 



SCIENCE. 



[N. S. Vol. XV. No. 370. 



tions are special and limited; but through 

 them we discover facts which may be 

 grouped with others of like kind. Presently 

 we shall reach the discrimination between 

 elements and compounds; and sooner or 

 later we shall find ourselves face to face 

 with one of the ultimate problems of all 

 science— the nature of matter itself. In 

 this problem all questions of chemical com- 

 position come to a focus ; it goes back of the 

 reaction to the substances which react; but 

 it belongs equally to physics, and its essen- 

 tial details admit of description only in 

 physical terms. Chemistry, however, is 

 doing the most towards its solution, for it 

 is through chemical researches that varia- 

 tions in the composition of matter are best 

 explained. The indebtedness of chemistry 

 to physics is thus fully repaid. 



"What is matter? Is it continuous or dis- 

 crete, atomic or made up of vortex rings in 

 the ether? These questions admit of only 

 partial answers, and doubtless their final so- 

 lution is unattainable by man. They are, 

 nevertheless, perfectly legitimate questions 

 for science to ask ; and a tentative reply, of 

 great practical value, is given by the atomic 

 theory. Whether it be true or false, 

 whether the chemical atoms are ultimate or 

 divisible, this doctrine is the connecting 

 thread upon which our profoundest gener- 

 alizations are strung, and it is hard to see 

 how we could do without it. Once a mere 

 speculation of philosophy, Dalton gave it 

 quantitative meaning ; and from his day to 

 the present every great advance in chemical 

 theory has found its clearest statement in 

 atomic terms. Chemical equations and 

 formula; the laws which correlate the 

 density of a gas with its composition; the 

 law of Dulong and Petit; our ideas of 

 valency and molecular structure; the 

 periodic law; and the relations of stereo- 

 chemistry, are all connected by the atomic 

 theory, whose retention in science is there- 

 fore fully justified. It may not be beyond 



criticism; indeed, it should be criticized; but 

 it would be the utmost folly to abandon the 

 theory before something better has been 

 framed to take its place. Vague and un- 

 satisfactory are the attempts which have so 

 far been made to supplant it. Physics, un- 

 aided by chemistry, may reach the concep- 

 tion of molecules; but the subdivision of 

 the latter, the identification of their parts, 

 is the function of the chemist alone. 



If the nature of matter is the first ele- 

 ment in the study of chemical reactions, 

 the nature of chemical union is the second. 

 If combination consists in a juxtaposition 

 of atoms, what is the force which draws 

 and holds them together ? Whether we can 

 answer this question or not, we may investi- 

 gate the laws under which chemical action 

 is operative, and so develop an important 

 portion of physical chemistry. Problems of 

 chemical equilibrium, of limiting condi- 

 tions, of aflinity and the speed of reactions, 

 all come under this heading, and these are 

 fit subjects for investigation in the labo- 

 ratory. For instance, chemical action is im- 

 possible at very low temperatures, and at 

 sufficiently high temperatures all com- 

 pounds dissociate ; each reaction, therefore, 

 is confined to a certain part of the ther- 

 mometric scale, which in many cases is 

 measurable. In other words, chemical 

 change is a function of temperature, no 

 matter what additional factors its complete 

 study may involve. It may also be effected 

 through the agency of electrical or actinic 

 impulses; and here again experimental re- 

 search has a wide field. Were physical 

 chemistry restricted, as it is not, to this 

 class of investigations alone, it would stiU 

 have abundant occupation. These illustra- 

 tions are enough for my immediate pur- 

 pose, but they could be multiplied indefi- 

 nitely. 



Directly growing out of these two funda- 

 mental questions, and partly identifiable 

 with them, are two other problems of great 



