and Unmeasured Wave-lengths. 395 



planets, leads us to think that we can best understand their 

 relations to solar energy by studying the atmosphere of our 

 own ; for our non-coinprehension of these relations is largely 

 due to our ignorance of certain physical data which have 

 never yet been obtained. 



While the general question for the physical astronomer, 

 then, is " What kind of transformation does the solar energy 

 suffer at the surface of any planet ? " we here seek a reply to 

 the simpler preliminary one, " What are the wave-lengths of 

 heat from non-luminous sources, such as the soil of this 

 planet ? " — a question which has never been answered, because 

 there have been no means of recognizing this heat when drawn 

 out into a spectrum ; indeed, we so habitually associate the 

 idea of a spectrum with that of light, that there is a certain 

 strangeness at first in the idea even of a "spectrum" 

 formed by a cold body like, for instance, ice. Yet the ice 

 surface must not only be capable of radiating heat to a still 

 colder body, but, according to our present conceptions of 

 radiant energy, be capable of giving a spectrum, whether we 

 can recognize it or not. It is the object of the present paper 

 to describe the actual formation of such spectra, and the 

 recognition of their heat in approximate terms of wave- 

 lengths. 



To distinguish between these new regions of research and 

 the older ones, let us briefly summarize our actual infor- 

 mation about wave-lengths, since on the latter the whole 

 question largely turns, and each extension of it, we may 

 agree, is a step toward an interpretation of everything about 

 the constitution of the universe which radiant energy may 

 have to tell us. Thus there is no exact relation known 

 between the periods of vibration of certain molecules in the 

 sun and the angles through which the rays announcing them 

 are refracted by a prism, while the wave-lengths of these 

 rays, if known, are capable of giving us quite other intelli- 

 gence. 



Yet our knowledge even of the wave-lengths of light is 

 comparatively recent, since it was only at the beginning of 

 this century that the labours of Thomas Young brought the 

 undulatory theory itself from the disfavour in which it had 

 lain ; and the memoirs in which Fraunhofer gave the first 

 relatively full and accurate measures of the wave-lengths of 

 light date no further back than 1814. 



The measures of Xewton, interpreted in terms of the 

 present theory, gave the length of the extreme violet waves 

 at 10>0 1 6 ^ 000 of an inch, and of the extreme red at i off 000 ? or in 

 millimetres 0*00042 millim. and 0"00067 millim. respec- 



