154 Britain's Heritage of Science 



remark that " heat throws the molecules of lime into such 

 a state of such rapid vibration that they become capable of 

 influencing the surrounding setherial medium and producing 

 in it the undulations of light." 



In 1845 William Allen Miller (1817-1870), Professor of 

 Chemistry at King's College, London, published some observa- 

 tions on flame spectra, which were not very accurate, and 

 his plates left it doubtful whether the bright bands or the 

 dark intervals between them ought to be looked upon as 

 the essential feature. This seems to have been one of the 

 stumbling-blocks of early investigators when comparing the 

 continuous spectra of ordinary flames with the discontinuous 

 spectra of incandescent substances. 



An important contribution to the subject was made by 

 William Swan (1818-1894), who, between 1859 and 1880, held 

 the Professorship of Natural Philosophy at St. Andrew's. 

 Swan was the first to introduce (1847) the collimator into 

 spectroscopic observations, and in 1857 he examined and 

 accurately mapped the spectrum of hydrocarbon flames. He 

 discussed the origin of the ubiquitous yellow line and came 

 to the correct conclusion that it is due to the presence of 

 minute quantities of sodium. 



The spectra of the electric sparks passing between poles 

 of different metals were first examined by Sir Charles 

 Wheatstone, and described in a communication to the British 

 Association in 1835. Unfortunately an abstract only was 

 published, but even the short account given ought to have 

 drawn attention to the extreme importance of the matter. 

 The spectrum of mercury was observed and accurately de- 

 scribed, and proved to be identical, whether the spark be taken 

 in air, oxygen gas, the vacuum obtained by an air pump, or 

 the Torricellian vacuum. From these observations the correct 

 inference was drawn that the spectrum is the result of the 

 volatilization and ignition (not combustion) of the ponderable 

 matter contained in the spark. The spectra of zinc, cadmium, 

 bismuth and lead were also obtained by taking the sparks 

 from poles of the melted metals. The paper was published 

 in full in the Chemical News in 1861, and was then found 

 to contain this significant passage : " the number, position, 

 and colour of these lines differ in each of the metals 



