164 



THE EXTREME INFRA-RED RADIATIONS. 



quite readily, and still more freely the precipitated chloride of silver. 

 The radiations peculiar to rock salt are completely untransmissible to 

 rock salt and chloride of silver, as well as to glass, gypsum, and fluor- 

 spar. Paraffin in layers of 1 millimeter thickness transmits nearly half 

 of these radiations of rock salt, and a comparison with the radiation 

 peculiar to sylvine shows paraffin becoming more and more trans- 

 parent as we proceed further in the infra-red. The same is true of the 

 transmitting power of quartz, fluorite, and gutta-percha. 



A sheet of isinglass transmits nearly two-thirds of each of these 

 kinds of radiation. The authors therefore employed this substance to 

 form a cell in which to put various liquids whose transparency Was to 

 be examined. Carbon-bisulphide and benzine were found to be very 

 transparent, petroleum somewhat less so, and toluene and xylene still 

 more opaque. Thus the fractions of the rays separated by rock salt 

 which traversed 1 millimeter thickness of carbon-bisulphide and xylene 

 were 98 rjer cent and 16 per cent, respectively. The case of olive oil is 



Oscillations electriques- 



i-t- 'I ■ ■ I 1 , I 1 1 J 1 1 Hr- 



11 II. I IS ' 16 ll 18 19 ' 20 ' 21 22 ' 23 \2> 



I I I Octaves J 



Longueurs d'or.de . 



Fig. 2. — The spectrum divided into octaves. The shaded parts have not been exposed. 



singular. While completely opaque to rays of the wave length 51/;, a 

 layer of it 1 millimeter thick transmits 20 per cent of those at 61,u. 

 Water, alcohol, and ether are completely opaque to both these kinds of 

 radiations, and while carbonic-acid vapor absorbs them but slightly, 

 water vapor absorbs them almost completely. 



These observations show us anew how marked are the differences in 

 the properties of ether waves of slightly different periods of oscillation. 

 It appears, according to modern theories, that if a substance is trans- 

 parent at one region of the spectrum, it is necessarily opaque at some 

 other. Indeed, if we except the case of the metals, we must search for 

 substances transparent in the infra-red among those opaque to visible 

 radiations. 



Nothing more clearly indicates the great generalization in the notion 

 of light that has taken place in the last decade than the simple statis- 

 tics of comparison between the extent of spectrum as now known and 

 the narrow visible stretch included between the violet and the red. 



This comparison becomes clearer in a graphic representation, which 



