306 Dr. T. Oamelley on the 



which research is most needed and in which there is most pro- 

 mise of interesting results. The application of this principle 

 "will also enable us to make predictions of phenomena still un- 

 known, and will at the same timo prevent many fruitless 

 researches. It is and will be, in fact, for some time to come 

 the finger-post of chemical science. 



Notwithstanding the importance of this subject, it has up to 

 the present been very much neglected. For though Mendcl- 

 jeff's results are well known, yet the details and the service 

 which his Periodic Law offers in the prosecution of chemical 

 research are far from being so. Were his memoir more ge- 

 nerally read, and the methods he proposes more widely applied, 

 many fruitless researches would be avoided, and many impor- 

 tant problems would be solved far more readily than is gene- 

 rally the case at present. 



Up to within the last few years the work of the chemist has 

 been very largely to collect facts. Now, however, the great ten- 

 dency is to generalize, connect together, and offer some expla- 

 nation of these facts. Thus, take a given element. Why should 

 it possess certain properties and another element certain other 

 properties ? And what connexion is there between the different 

 properties of the same element ? What, for instance, consti- 

 tutes the true difference between Ag and CI, and why are they 

 different ? Why should silver melt only at 1000° C. and be a 

 heavy solid at the ordinary temperature, whilst chlorine is a 

 gas, very much less dense than silver? Why should silver 

 have a great affinity for bromine, whilst chlorine has but very 

 slight affinity for it ? The greatest and highest object of the 

 chemist and physicist, therefore, is to endeavour to offer some 

 explanation of these and of a host of similar phenomena, and 

 to get as near as possible at the root of the matter. In 1864 

 Newlands made the first great step in advance, which advance 

 was increased and placed on a firmer basis by Mendeljeff in 

 1869. 



The stage at which we have now arrived, and at which work 

 is still being carried on with ever increasing activity, is the 

 tracing of the interconnexion of properties. For this purpose, 

 what w t c have to do is to find what properties vary regularly 

 with other properties. A well-known example of this is Du- 

 long and Petit' s law of " Specific Heats," according to w T hich 

 the specific heat of an element is inversely as its atomic weight; 

 also Gay-Lussac's law of vapour-density, which states that 

 the vapour-density of a gas = \ its molecular weight. Also, 

 I have recently shown (Dent, cliem. Ges. Ber. xii. p. 440, 

 March 1879) that the greater the coefficient of expansion of 

 the elements by heat the lower the melting-point. And more 



