544 THE PRINCIPLES OF SCIENCE. [CHAP. 



Sir W. Thomson was enabled to anticipate the following 

 curious effect, namely, that an electric current passing in 

 an iron bar from a hot to a cold part produces a cooling 

 effect, but in a copper bar the effect is exactly opposite in 

 character, that is, the bar becomes heated. 1 The action 

 of crystals with regard to heat and electricity was partly 

 foreseen on the grounds of theory by Poisson. 



Chemistry, although to a great extent an empirical 

 science, has not been without prophetic triumphs. The 

 existence of the metals potassium and sodium was fore- 

 seen by Lavoisier, and their elimination by Davy was one 

 of the chief experimental, cnwis which established Lavoi- 

 sier's system. The existence of many other metals 

 which eye had never seen was a natural inference, and 

 theory has not been at fault. In the above cases the 

 compounds of the metal were well known, and it was 

 the result of decomposition that was foretold. The dis- 

 covery in 1876 of the metal gallium is peculiarly inter- 

 esting because the existence of this metal, previously 

 wholly unknown, had been inferred from theoretical con- 

 siderations by M. Mendelief, and some of its properties 

 had been correctly predicted. No sooner, too, had a 

 theory of organic compounds been conceived by Professor 

 A. W. Williamson than he foretold the formation of a 

 complex substance consisting of water in which both 

 atoms of hydrogen are replaced by atoms of acetyle. 

 This substance, known as the acetic anhydride, was after- 

 wards produced by Gerhardt. In the subsequent progress 

 of organic chemistry occurrences of this kind have become 

 common. The theoretical chemist by the classification of 

 his specimens and the manipulation of his formulae can 

 plan out whole series of unknown oils, acids, and alcohols, 

 just as a designer might draw out a multitude of patterns. 

 Professor Cayley has even calculated for certain cases the 

 possible numbers of chemical compounds. 2 The formation 

 of many such substances is a matter of course ; but there 

 is an interesting prediction given by Hofmann, concerning 

 the possible existence of new compounds of sulphur and 



1 Tait's Thermodynamics, p. 77. 



2 On the Analytical Forms called Trees, with Application to the 

 Theory of Chemical Combinations. Report of the British Association, 

 1875, P- 257- 



