Physical Constitution of the Sun. 303 



tered and rendered incandescent by the spark is very small in 

 proportion to the whole space within which the passage of the 

 spark takes place (as it was in Wiiilner's more recent experi- 

 ments, in which the tubes employed possessed throughout a 

 clear width of about 2 centims., and the electrodes were dis- 

 tant 7*5 centims. from each other*), the extent of gas made 

 incandescent will, under an approximately constant pressure, 

 expand, and, in correspondence with Gay-Lussac's law, dimi- 

 nish its density. From this it is evident that for the estima- 

 tion of the quantity of gas which lies on the visual line passed 

 through perpendicularly by the spark -track, the knowledge of 

 three quantities is required, viz. : — 



1. The original density of the gas within the tube; 



2. The thickness of the spark-track ; 



3. The temperature of the spark. 



The first of these can be determined, the second approximately 

 estimated for the present purpose ; the third remains unknown. 

 Yet, considering that, the condition for the comparability of the 

 terrestrial hydrogen-spectrum and that of the chromosphere rests 

 on the assumption that the physical proportions are substanti- 

 ally the same in both phenomena, the temperature deduced above 

 for the chromosphere must, by way of approximation, be pre- 

 supposed also for the spark-track in the Geissler's tube. Lastly, 

 the densities of the gas during and before incandescence would 

 then be inversely as the absolute temperatures in the two states. 

 Taking for granted, then, the temperature-values obtained above 

 for the chromosphere, if the temperature of the hydrogen in the 

 Geissler's tube before incandescence equals that of melting ice, 

 the density of the glowing spark-track would be about % oo °^ 

 that of the gas contained in the tube at 0° C. 



If it be further assumed that the thickness of the spark-track 

 amounts to about 1 millim., the metre having been employed 

 as the unit of length in the above formula?, the density of a 

 quantity of incandescent hydrogen which in the unit of volume 

 contains the same mass as an extent passing rectilinearly right 

 through the chromosphere with the unit of surface as cross sec- 

 tion, will be only ^oo^odd °^ tne density of the hydrogen at 

 0° C. in the Geissler's tube. Hence, if a denotes the original 

 density, it is reduced in the spark- track of the assumed thick- 

 ness to ^r, and must be further diminished to the thousandth 



part of this quantity if expanded from the assumed extent of 

 millimetre to that of a metre as the unit of length. The value 



* Pogg. Ann. vol. cxlvii. p, 325. 



