Clifton College Scientific Society. 89 



and importance, and it has been found that different compounds 

 of this element, when raised to the state of incandescent gas, give 

 somewhat different spectra. For exa,mple, cyanogen (CN) yields 

 several lines which common coal-gas does not, among which are 

 several blue and red. The knowledge of the carbon spectrum has 

 been put to good use in the manufacture of steel by the Bessemer 

 process. Steel differs from cast-iron in containing less carbon, 

 and this is actually burnt out by allowing a stream of air to pass 

 through the cast-iron when in a molten state. The flame which is 

 given oft' is peculiar, and, by observing it with the spectroscope, we 

 can tell the moment at which the operation must be stopped. The 

 exact point at which this must be done is when the carbon lines 

 disappear and the spectrum becomes continuous. If we allow it 

 ■ to go on a little longer than is proper, we shall obtain the steel so 

 viscid that it cannot be poured out of the ladle ; if, on the other 

 hand, we stop too soon, the iron will contain so much carbon that 

 it will be too brittle for use. Among the substances in the Besse- 

 mer flame, sodium, potassium, silicon, iron, carbon, hydrogen, and 

 nitrogen have been detected. 



It has long been known that certain bodies have, at the ordin- 

 ary temperature, the power of selecting a certain kind of light and 

 absorbing it. This is called selective absorption. For example, 

 when lohite light passes through the violet vapours of iodine, 

 or red nitrous fumes, dark bands will be seen. Some coloured 

 gases do not give any bands, while other colourless gases give 

 spectra full of dark lines. The solutions of many salts give very 

 peculiar bands, especially didymium, which can be detected in this 

 way with the greatest ease. Blood also gives very curious bands. 

 Cruorine gives quite a different absorption spectrum from hsematin. 

 And there is a very interesting fact connected with this subject. 

 Poisoning by CO2 can be readily detected, owing to the different 

 spectra which pure blood and that containing carbonic dioxide 

 produce. Dr Herapath of Bristol, in concluding a pamphlet 

 entitled, ' On the Use of the Spectroscope and Microspectroscope 

 in the Discovery of Blood-stains and Dissolved Blood, and in 

 Pathological Inquiries,' says, ' Although spectrum analysis does 

 not go one step farther than we were before in our powers of dis- 

 criminating human blood from that of other mammalian or red- 

 blood creatures, yet it gives us greater facilities of demonstrating 

 the presence of the colouring matter of blood, even in inconceiv- 

 ably minute and almost invisible proportions, whilst the facUity 

 with which the observations are made is a great, if not the greatest, 

 recommendation to the employment of this method whenever 

 practicable.' 



It is a well-known fact in optics that some substances, as the 



