the Spectra of Carbon and Hydrocarbon. 557 



In the attempts to photograph these lines, the fainter 

 members of the other series starting from the head of the 

 main bund were being simultaneously recorded on the 

 plates. The enlargement reproduced reveals the pairs of 

 lines proceeding with gradually increasing separation up 

 to what appear to be two nebulous pairs or narrow bands. 

 Between these and the new faint series only a short space 

 intervenes, so that the seemingly void region has been 

 greatly reduced. 



In structure the faint series closely resembles the main 

 series of the tail, but its scale is smaller. 



As this paper is not intended to be a detailed communi- 

 cation, no discussion of these line series is attempted and 

 their investigation awaits more exact data. 



Some Bands of Unknown Origin associated with the 

 " Swan " SpectriLm. 



In the course of some work on the Meeker flame spectrum, 

 some small-scale spectrograms were taken with a grating 

 camera. 



On close examination of these, it was noticed that certain 

 faint bands were visible in the region between the green 

 and the blue carbon flutings. These bands were too faintly 

 recorded to enable anything more than rough wave-lengths 

 to be obtained, nor was any structure visible in them; but 

 during the investigation of the lines of the hydroearben 

 bands with the prismatic spectrograph, the new bands 

 appeared on the negatives with much greater clearness. ]t 

 became important, then, to measure them more accurately 

 and endeavour to ascertain something of their structure. 

 ' At best the bands are very feeble, being only just visible 

 on the faint continuous spectrum which extends between the 

 " Swan " bands. Ordinary plates were used, as the range 

 of wave-length under investigation fell within their limits 

 of sensitivity ; but long exposures were necessary owing to 

 the fain til ess of the light. As in the case of the hydrocarbon 

 bands, a limit to successful photography was imposed by \he 

 fogging produced by the feeble continuous radiation; but 

 in spite of this difficulty, some useful spectrograms were 

 obtained. 



For the purpose of wave-length determinations the spark 

 spectrum of iron was employed, the discharge being passed 

 between iron electrodes with capacity and self-induction 

 in circuit. It fortunately happens that iron gives excellent 

 reference lines in the region under investigation, and a 



