NA TURE 



49 



THURSDAY, MAY i8, 



OSTWALD'S GENERAL CHEMISTRY. 

 I.eJirbuch der Allgemeinen Chemie. Von Dr. Wilhelm 

 Ostwald. Band I. 1891, Band II. Theil I. 1893. Zweite 

 Auflage. (Leipzig : Wilhelm Englemann.) 



THE conception of molecule is essential in explaining 

 the phenomena of both chemistry and physics. 

 ( Porosity and compressibility point to the conclusion that 

 1 matter does not entirely fill space, to account for the 

 dispersion of light requires that matter should have a 

 I grained structure ; these and countless other physical 

 facts find an explanation in the conception of molecule. 

 ' Moreover, from various observations, more especially on 

 the properties of gases and the phenomena of surface 

 'tension, the size of molecules can be approximately cal- 

 iculated, and in terms of the idea of molecule deduced in 

 ways such as these physical properties are explained. 



The chemist, on the other hand, has arrived at the 

 need of the conception of molecule from totally different 

 considerations. In the early days of his science, when 

 the laws of combining proportions and of chemical 

 equivalents were taking definite shape, the revival of the 

 conception of atom was of immediate service in further- 

 ing the progress of chemistry. It was not long in 

 becoming apparent, however, that the conception of 

 atom alone was insufficient to meet the facts. 



The relative numbers of atoms entering into the com- 

 position of compounds was a matter of doubt until 

 .Avogadro's hypothesis was accepted, and until it was 

 granted that definite groups of atoms — chemical mole- 

 cules— were concerned in chemical processes. 



The chemist has thus built up his conception of mole- 

 cule in accordance with chemical facts ; he regards it 

 IS a structure composed of parts, and in order to explain 

 he existence of isomers, he has to assume definite rela- 

 tive arrangements of the atoms within a molecule. 



From the fact that the two conceptions of molecule 

 lave been derived independently of one another, it has 

 come about that physical properties are discussed more 

 )r less apart from the chemical nature of the substances 

 ;xamined, and for this reason within recent times there 

 las arisen a fascinating field of inquiry on the border- 

 and of chemistry and physics. For it has been urged, 

 ' Is it not possible to trace the cause of physical pheno- 

 nena beyond the physical molecule ? " If, as the chemist 

 las shown, the molecule is a structure composed of 

 )arts, is it not possible that these parts of molecules are 

 he units to be dealt with.' In short, " Is not the ulti- 

 nate cause of physical as well as of chemical phenomena 

 o be ascribed to the chemical atoms and their mutual 

 elationships .' " 



Already this question has been answered in several 

 vays, and in none more striking than by those investiga- 

 ions which are concerned with the physical constants of 

 ubstances and their chemical nature. Here it has been 

 hown that the magnitudes of many physical constants 

 re conditioned by the nature, number, and arrangement 

 if the atoms which compose molecules and that fre- 

 [uently definite changes in chemical nature bring about 

 NO. 1229, VOL. 48] 



definite quantitative changes in the magnitude of physical 

 constants. 



Books dealing with such investigations as these are 

 but few, indeed the first volume of the book before us is 

 practically the only one which gives a comprehensive 

 view of what has been done in this direction. If we ex- 

 clude those parts which are purely physical and which 

 are concerned with familiarising the reader with the 

 physical properties to be treated, the volume may in 

 the main be taken as linking on the chemical to the 

 physical conception of molecule, in so far as to show that 

 the magnitudes of physical constants are functions of 

 molecular weight and molecular structure. 



Thegeneral arrangement of the contents of this volume is- 

 pretty much as it was in the previous edition, although very 

 few pages remain as they were, and the introduction of 

 recent investigations has increased the size of the volume 

 by about one-third. The atomic hypothesis and the laws 

 upon which it is based are first treated, then follows a use- 

 ful summary of the various atomic weight estimations,, 

 from which are deduced the probable values of those funda- 

 mental constants,values which are already finding their way 

 into current literature. The numerical relations existing; 

 between the atomic weights of the elements constitute 

 the concluding portion of this the first of the six books 

 into which vol. i. is divided. Succeeding books deal re- 

 spectively with the physical properties of gases, liquids,, 

 solutions, and solids, and with the relations existing be- 

 tween the physical properties and the chemical nature of 

 the substances. 



Solutions are, in this edition, for the first time treated 

 in a separate book, which with certain additions has beea 

 translated into English by Mr. Pattison Muir, and has 

 already been noticed in these columns (Nature, vol. xlv. 

 p. 193). Electric conductivity and electrolysis now find 

 a place in vol. ii. under electro-chemistry. The sixth and 

 last book of vol. i. deals with chemical systematics — the 

 criteria by which atomic weights are chosen, the 

 periodic law and the relations between the physical con- 

 stants of the elements and their atomic weights, and the 

 molecular theory and the structure of chemical com- 

 pounds in which the doctrines of valency, isomerism, &C.,. 

 are discussed. 



The peculiar interest which attaches to connections- 

 between the physical constants of substances and their 

 chemical nature lies in the fact that an idea is thereby 

 obtained of the constitution of the substances as they 

 actually exist. Structure as deduced from purely chemical 

 methods is founded upon reaction. The compound has 

 to be decomposed before its constitution can be deter- 

 mined, and occasionally such methods lead to ambiguous 

 results. Examples are steadily multiplying of compounds 

 which in one reaction appear to correspond with one 

 formula, while in another reaction a different formula 

 better represents their chemical behaviour. Already 

 measurements of physical constants have been applied 

 to some such cases and have served to indicate that the 

 structure of a pure substance may be conditioned by its 

 temperature. At high temperatures, for example, acetyl 

 acetone would appear to exist in the ketonic condition, 

 CH3CO.CH.i.CO.CH3, while as temperature falls it would 

 seem as if a gradual transition to the alcohoUc conditions, 



D 



