561- REFOKT — 1902. 



It is then only as colligated equalities, established by experiment, that gaseous 

 volumes, osmotic pressures, and other properties of substances come into considera- 

 tion, first as enforcing the truth of the conception of the indicated quantities as 

 equal, and then as the means of molecular measurement without resort to 

 chemical change. But of the purposes served by the coUigative properties, that 

 of giving molecular measurements without recourse to the evidence afforded by 

 chemical change is well known to be of the very widest application. To 

 determine chemically the molecular equalities of substances, single chemical 

 changes of suitable character, clianges which are cases of double decomposition, 

 have to be looked for ; and to know these with the desirable degree of certainty 

 calls for a much larger acquaintance with the chemical behaviour of the substances 

 than can usually be gained at the early stage of work when the knowledge of the 

 molecule is of the utmost assistance in the further investigation of the nature of 

 the substances. Consequently, it is nearly always through recourse to physical 

 methods that the molecule is first ascertained, and then through the molecule the 

 certainty acquired that some particular interaction is a single one, thus reversing 

 the normal order of things, which undoubtedly is that the molecule in chemistry, 

 however it may have been tirst determined, is recognised as such by being what 

 it is in chemical change. 



I shall have been wholly misunderstood by you if you suppose that I would 

 make light of the importance of the balance in chemical operations, or of the 

 value of its indications in chemical investigations. Once the weights of mole- 

 cular or atomic quantities have been ascertained the balance becomes the most 

 accurate and generally the most easily applied instrument for apportioning 

 substances in these quantities. Chemical interaction, to be employed in this way 

 and without the aid of the balance, is practically useless, for the reason that it 

 involves the destruction of the quantities it measures. Out of this dependence on 

 the balance arises the exceeding importance of accurate tables of atomic weights, 

 from which molecular weights are derived by addition ; but the place for these 

 tables is not on the walls of the lecture-theatre, but in the laboratory pocket-book 

 and, perhaps, in the balance-room. Besides the use of the balance and of atomic 

 weight tables for getting and calculating out molecules of different substances at 

 pleasure, there is the indispensable service they perform in enabling chemical 

 analysis to be carried out and applied to the solution of the problems offered by 

 chemical change. The primary problem of every science is to find some element 

 of sameness in the diversity of its phenomena, in order that they may be com- 

 pared, a problem which was solved for chemistry to a l.irge extent by Dalton, and 

 ceased to exist wheu the distinction had been made between molecule and atom. 

 But this having been solved, there comes the other problem, namely, to find 

 definite, that is, quantitative differences in the midst of the uniformities, and these 

 for the chemist are differences of mass or weight. Through that redistribution of 

 mass which attends chemical interactions it has been possible to trace out to some 

 extent the nature of the transformation of substances and develop the science on 

 the lines of chemical composition and chemical constitution. Thus, then, the 

 balance has become and will continue to be the necessary instrument of chemical 

 research ; but again I would remind you that it records its facts in units which 

 are not ours, and of which we avail ourselves only as the means to au end. 

 Sodium chloride is chemically composed, not of 3545 equal parts of chlorine with 

 2305 of the same equal parts of sodium, but of equal quantities of these simple 



The theory of chemical molecules or equalities and their relations to the 

 equalities between the weights and gaseous volumes of different substances wore 

 brought to light not by Richter's law of chemical combining proportion*, and not 

 hv A"vo"'adro's hypothesis as to there being equal numbers of particles in the same 

 volume "of different gases, but in the first place by Dalton's atomic theory and 

 Gay-Lussac's law of simply related ga.«eous volumes in chemical change ; and 

 then much more fully in the middle of the last century, thi'ough the brilliant 

 work of Gerhardt, Williamson, Laurent, Odling, AVurtz, and others, in the purely 

 chemical field. Dalton gave us the couceptiou of the molecule, though confused 



