144 



SCIENCE 



[N. S. Vol. XXVI. No. 657 



The author attempts to represent the afEmities 

 of the elements for each other by means of a 

 diagram. This diagram for any particular 

 element is constructed by arranging the ele- 

 ments according to the periodic table. Then 

 over each element erect a perpendicular 

 whose length shall represent the affinity of the 

 chosen element for the element in the table. 

 Joining the points (ends of the perpendicu- 

 lars) thus obtained, you have, a surface vifhich 

 represents at a glance the affinity of the 

 chosen element for all of the elements. 



The fact that such a surface could be con- 

 structed by any one having a knowledge of 

 the proper lengths of the perpendiculars needs 

 no proof, and the author's mathematics on this 

 subject could, we think, have been omitted 

 with advantage. The real difficulty arises in 

 determining the leng-th of the perpendicular 

 which shall represent the affinity in any given 

 case. Where the heat of formation is large this 

 is taken as measuring the attractive force. 

 Where the heat of formation is small, or the data 

 insufficient, the author determines the proper 

 length by a comparative study of the com- 

 pound with regard to its stability, etc. The 

 surfaces as given by the author for thirty-one 

 substances are therefore not claimed to be ex- 

 actly quantitative, but only qualitative. The 

 author points out the readily perceived fact 

 that the form of the surface would be altered 

 should the arrangement of the elements be 

 changed. Also it would be altered by changes 

 in pressure and temperature and would 

 change its form completely with a change in 

 the valence of the element. 



We think it hardly probable, even where the 

 heat of formation is large, that it gives any 

 very good idea of the actual size of the at- 

 tractive forces involved. For the elements 

 reacting are usually previously not mona- 

 tomic and the heat of formation is the sum of 

 the several reactions involved, to say nothing 

 of a variable amount due to the physical 

 changes. Moreover, T. W. Richards and J. 

 Traube have separately pointed out that the 

 heat of formation is probably largely depend- 

 ent upon the changes in volume suffered by 

 the reacting elements. When these facts are 



all considered the affinity surfaces constructed 

 by the author at the expense of so much labor 

 can hardly be taken as more than a guess. 

 Much wiU have to be learned before any true 

 affinity surface can be constructed. 



From various facts cited by the author as 

 collateral evidence bearing upon his point of 

 view some of his conclusions are worthy of 

 note either as being new or because they are 

 derived from a more or less novel standpoint. 

 Thus on page 42 : 



It seems, in fact, that the force exerted by the 

 bromine atom is always less than the force ex- 

 erted by the chlorine atom on a given element or 

 radicle, but is always proportional to it. 



The chemical attractions which A exerts on the 

 various elementary atoms or radicles must be 

 either equal or proportional to the chemical at- 

 tractions which B exerts on the same atoms or 

 radicles. 



On page 74 : 



It is the intensity of internal atomic forces 

 with which the atoms are attracted together in the 

 molecule which determines the intensity of the 

 external attractive force between molecule and 

 molecule, and therefore the volatility of the com- 

 pound. 



On page 99: 



Strangely enough, the melting-points of similar 

 compounds does not appear to depend upon the 

 weight of the molecules, but upon their chemical 

 nature, because, as a rule, they melt at tempera- 

 tures which lie close together in spite of great 

 differences in their molecular weights. 



On page 101: 



The melting-points and solubility in water can 

 not depend largely upon the magnitude of the 

 molecular weight, but must depend upon the chem- 

 ical forces. 



On page 111 : 



Consequently we infer that the volatility of 

 compounds is determined almost entirely by the 

 intensity of the forces which bind together the 

 atoms in the molecule. 



On page 219 : 



If this be so, a non-metal is nothing more or 

 less than a substance viewed at a temperature too 

 low for it to assume metallic properties; and con- 

 versely, a metal is a substance viewed at a tern- 



