in the Ultra-red Region of the Spectrum of the Sun. 239 



ground smooth and closed with plates of rock-salt of 2 millim. 

 thickness. It being very inconvenient and difficult to keep 

 the faces of the rock-salt clean and clear for any length of 

 time, this way of filtering the rays is troublesome and expen- 

 sive. Subsequently, therefore, instead of the iodine solution, 

 I used a perfectly black plate of ebonite of 0'3 millim. thick- 

 ness, this substance being, according to Abney and Festing*, 

 very diathermanous for rays of great wave-length. This plate 

 showed itself in fact very transmissive also for the extreme 

 ultra-red rays; and its employment is very convenient. 



In our arrangement all the reflections took place at metallic 

 surfaces; and any passing of the rays through glass was com- 

 pletely avoided till they reached the very thin glass bulb of 

 the torsion- apparatus, through which the rays had to pass 

 before meeting the radiometer-vane. As, however, any ab- 

 sorption of rays by this glass acts upon the vane by producing 

 motion in the same direction as the absorption by the lamp- 

 blackf , even the rays which may have been absorbed by the 

 glass bulb were also effective. 



Although the superposition of the luminous rays of the 

 second and the obscure rays of the first spectrum affords a 

 convenient means for approximately estimating the wave- 

 length of the obscure rays occurring in each place, I arranged 

 the setting-up of the apparatus so that an exact measurement 

 of the wave-length was possible. For this purpose I placed 

 the diffraction-grating G on a spectrometer-stand, movable 

 round a vertical axis, from which the telescopes were removed. 

 The position of the stand could be read on the orientation- 

 circle, which was divided into intervals of five minutes of arc, 

 by means of a microscope fixed by an arm to the stand, and 

 the thread cross of which could be displaced laterally with the 

 aid of a micrometer-screw, so as to permit whole minutes to 

 be read off directly. The spectrometer had three adjusting- 

 screws in its base ; further, the stand stood separately upon 

 three screws ; and, lastly, the mirror Gr was placed in a box 

 which, again, rested on three adjusting-screws ; so that the 

 stand could be shifted at discretion with respect to the micro- 

 meter, and the level of the mirror with respect to that of the 

 stand. 



II. Method. 



As in our arrangement the grating only can be rotated, 

 while all the other parts of the apparatus are fixed, it was 

 necessary to employ a somewhat different way of measuring 



* Ahney and Festing, Chemical News, xliii. pp. 176, 177 (1881). 

 t Conf. Pringsheim, Phil. Mag. February 1883, pp. 105-110. 



