186 AIR. E. C. C. BALY ON THE SPECTRA OF NEON, KRYPTON AND XENON. 
Leyden jar and a spark gap are placed in the circuit no decided change takes place m 
either the appearance or the spectrum of the discharge. Krypton and xenon, on the 
other hand, resemble argon in this respect that they both possess two spectra, one 
l)eing obtained when the direct discharge is passed and the other when a Leyden jar 
and spark gap are placed, in the circuit. The jar and spark-gap spectrum, or the 
second spectrum as I have called it in the tables below, is veiy much moie complex 
than the first spectrum, wherein an analogy is to he found with the two spectra of 
argon. When the discharge is passed through krypton without the use of a jar the 
spectrum obtained consists of a few lines, by far the most important being the yellow 
and green lines and also a group in the blue. As these lines have about the same 
visual intensity, the colour of the discharge is rather nondescript and appears to he 
different to different observers. The jar and spark discharge, on the othei hand, 
presents a fine sky-blue colour and its spectrum contains a considerable number of 
lines chiefly in the blue. The simple discharge through xenon is not "veij biilhant 
and is bluish in colour, being characterised by a group of blue lines less refrangible 
than the group in the first krypton spectrum. The second xenon spectrum is even 
more complex than the second krypton spectrum ; it is characterised by ceitain blight 
green lines which determine the colour of the discharge. 
The time of exposure necessary for photographing the spectra depended of course 
upon the brightness of the discharge; while in the case of the second xenon spectium 
an exposure of two to three hours was sufficient, the lines of the fiist spectium were 
so weak that an exposure of twenty-four hours was often required. The v a\ e-lengths 
of the lines were all determined liy interpolation between the lines of the arc spectrum 
of iron, and the wave-lengths of the principal lines in this spectrum as determined by 
Kayser were adopted as standards. Many photographs were taken of the tv o 
spectra superposed upon one another, caTe being taken that they were both correct!} 
placed in relation to each other. The wave-lengths of as many lines as possible veie 
determined in this way from several plates, and the mean values of these weie 
employed as standards in the measurements of several series of photographs taken of 
the new spectrum alone. In this way the wave-lengths of the fainter lines were 
obtained and also a certain amount of check was applied to the measurements of the 
standards themselves. Excepting a few lines in the red region and the fainter lines, 
nearly every line was measured in two orders, which ensures a correct i elation 
between the different regions of the spectrum. 
It was stated above that the spectra of the new gases are composed of bright and 
well defined lines ; it will he seen, however, from the tables, that certain lines aie 
marked as lieing diffused, which may appear at first sight to be rather contradictoi}. 
The explanation is to he found in the fact that the distinctness of the spectium and 
the definition of the lines depend to a great extent upon the pressure of the gas in 
the vacuum tube. If this pressure exceeds a certain small amount, the whole 
appearance of the spectrum is altered, for it tends to become confused, and at times 
