and on those produced by the Earth's Atmosphere. 523 



of mica, and thick gloves, and placing the tubes in cylinders of 

 tinned iron with narrow slits to admit the light, there is little 

 danger of any serious accident. 



When the gas is in the liquid state, it produces none of the 

 fixed lines which I have described, and exercises no other action 

 upon the spectrum than any ordinary fluid of the same orange 

 colour. 



In examining the structure of the solar spectrum, FKAUNHO- 

 FER seems to have put forth all his strength in determining the 

 position of the principal lines, A, B, C, D, E, b, F, G and H,* 

 which he had selected as equidistant as possible, for the purpose 

 of measuring their angular distances in different media, and thus 

 obtaining the most accurate data for the construction of the 

 achromatic telescope. These measures he has given with the 

 greatest exactness for various kinds of crown and flint glass and 

 for a few fluids, and he has thus put it in the power of the practical 

 optician to construct achromatic object-glasses, with a degree of 

 certainty and perfection hitherto unknown. 



This method, however, notwithstanding its high value, is 

 not easily applicable in practice, and from the nice observations 

 which it involves, we have reason to believe that it has not been 

 used by any other artist than FRAUNHOFER himself. The diffi- 

 culty of procuring out of the mass of glass to be employed, prisms 

 sufficiently pure to show such narrow lines as E, or the two which 

 constitute D f, of obtaining the sun when his light is wanted, 

 and of observing and measuring the distances of the fixed lines 

 in a spectrum constantly in motion, are insurmountable obstacles 

 to the general adoption of so refined a method of measuring dis- 

 persive powers. 



* Six of these, viz. B, D, b, F, G, and H, were discovered by Dr WOLLASTON. 



f These lines are also the most important, as the most luminous part of the spec- 

 trum lies between them. 



VOL. XII. PART II. 3 X 



