42 Prof. Liveing, On the variation of intensity of [Nov. 28, 



and also widened, the more diffuse being most widened. The 

 narrow, sharply defined, band about A 427 retains its character, but 

 is nevertheless sensibly widened. 



2. The difference between the nitrate and chloride is that the 

 light on the more refrangible side falls away in the spectrum of 

 the chloride, from about A 420, at a more rapid rate than in the 

 spectrum of the nitrate. It is as if a broad, weak, absorption 

 beginning from the more refrangible part, gradually extended down 

 to the violet region, in the chloride, but not in the nitrate. 



Also the light which gets through the chloride in the green 

 seems a little less than that which gets through the nitrate, as 

 if the chloride had a faint, broad, absorption band in the green 

 which the nitrate had not. 



If we accept the theory that the absorptions common to the 

 salts in dilute solutions are due to their common ions, namely, the 

 didymium and other metallic ions, then these broad absorptions, 

 which appear to be added when a stronger solution of the chloride 

 is employed, may be ascribed to the undecomposed chloride, or 

 they may possibly be due to the chlor-ion. As they have not 

 hitherto been observed to be produced by a solution of hydrochloric 

 acid, which must contain the chlor-ion, it is not probable that they 

 are due to the chlor-ion. 



Turning to the more refrangible section of the spectrum we 

 find, with the weakest solutions (g^th the original strength), 

 chloride and nitrate, four well-marked absorption bands between 

 lines N and of the solar spectrum, which may reasonably be 

 ascribed to the didymium and other metallic ions. Light passing 

 through the nitrate begins to fall away a little above (A 344), 

 and thence onwards diminishes so rapidly that beyond about A 333 

 nothing is seen in the photograph ; while with the chloride light 

 comes through quite to the edge of the plate (A. 315) but with 

 diminishing intensity from A, 345 onwards. 



With stronger solutions, -^ the original strength, all the bands 

 are expanded, and in the chloride a broad, diffuse, band shews 

 itself between A 333 and A 326, and a faint, very diffuse band 

 about A 338. The nitrate lets hardly any light through beyond 

 A 338, and the light which gets through about A 344 is very much 

 weakened. 



With still stronger solutions all light above A 360, or thereabouts, 

 is absorbed by the nitrate ; also more and more is absorbed by the 

 chloride, so that with the strongest solutions nothing at all is seen 

 in the plates beyond A 370. 



In fact the fading of the light in the case of the chloride seems 

 due to the gradual increase of a very wide and diffuse absorption 

 extending farther and farther towards the less refrangible side. 



