6 S. P. Langley — Observations on Invisible ILnt-Sprdrn.. 



trum of a surface below the freezing point of water. The 

 Leslie cube used in these experiments was filled either with a 

 freezing mixture, or with water kept gently boiling by a Bun- 

 sen burner underneath ; or again, when measurements from a 

 source at an exactly determinable higher temperature were de- 

 sired, with aniline, which has a boiling point of about 178° C. 

 A condensing apparatus connected with the cube, in the latter 

 case prevented the escape of the aniline vapor. It was also 

 found possible to keep the cube at any intermediate tempera- 

 ture within sufficiently narrow limits bj^ properly adjusting 

 the flame. 



The apparatus is shown in Plate 2.' Between the blackened 

 side of the' Leslie cube C and the spectrometer slit S, were in- 

 terposed a large pasteboard screen (a) and a flat copper vessel 

 (b) filled with broken ice, bolh pierced with apertures slightly 

 larger than the slit, to allow the passage of the rays, and the 

 exposures were made by withdrawing a third hollow screen (c) 

 made of copper and filled with ice, which cut off the radiation 

 of the cube from the slit when it was in place. 



The train for forming the spectrum upon the bolometer face 

 consisted of two rock-salt lenses L, L, and the rock-salt prism P. 

 Each lens is 75 mm in diameter and SoO 111111 focus for visible rays. 

 For the infra-red rays measured on, the focus is from one to two 

 centimeters greater than this. The prism is made from an un- 

 usually perfect piece of rock-salt, and is 64 rrun on a side. Its 

 constants having been fully given in this Journal for December, 

 we will only repeat here for convenience that its refracting 

 angle is 59° 57' 54'', and that the indices of refraction for the 

 Fraunhofer lines H 2 , b„ A, are 1-56920, 1-54975, 1-53670, re- 

 spectively, while that of 12, the farthest considerable absorption 

 band in the infra-red of the solar heat spectrum is l - 5268, cor- 

 responding to a known wave-length* of l/*-82. With this train, 

 composed entirely of rock-salt, and an ordinary eye-piece, the 

 Fraunhofer lines are very distinctly visible in either sunlight 

 or moonlight. The lenses, prism, slit and other parts of the 

 train were mounted upon the large spectrometer (described in 

 this Journal, xxv, 1883, and in the Mt. Whitney Eeport, 

 Chapter xi). 



To illustrate the use of the apparatus, we give below in de- 

 tail the observations of March 20, 1885, for determining the 

 form of the energy curve in the spectrum, of a Leslie cube at 

 178° C. The temperature of the room was —7° C, so that 

 the excess of the temperature of the cube over that of the 

 room was 185° C. 



The reading of the circle was made 0° 0' 0" when the spec- 

 trometer arms were in line, and the direct image of the slit fell 



* Given by a misprint 1^-32 in the December number of this Journal. 



