xx PROCEEDINGS. 
solution, purple in color, absorbs certain light, colored glasses absorb light, 
etc., but all manganese salts, colored or uncolored, do not absorb light, 
nor if the solutions are of different colors, though of the same element, 
do they absorb the same portions of light. | Gases absorb light, and on 
looking towards the sun with a spectroscope, faint bands are to be seen 
in the spectrum. Certain metals when heated, give certain varieties of 
light as, Na yellow, potassium blue, copper green, etc.,and these same 
substances, converted into vapour, absorb the same light that in their 
white hot condition they emitted, but here we have the phenomenon of 
a colorless solution acting like a colored solution or like a gas. This of 
course gave a test for Di. as soon as the absorption bands were mapped 
and thoroughly defined. This to the rare earth chemist was a valuable 
discovery, out in its application it calls forth the highest skill of the 
chemist, for he must be able to fix upon the exact bands and say this is 
given by Di., this by Er., ete. Here is displayed a fresh peculiarity of 
the rare earths, for while there are no elements outside these which 
give bands, there are several within the group which do, viz.,— 
Di. gives 31 well defined bands, Sm. 12 not so well defined (and 
some of them disputed, the subject is being thoroughly investigated by 
Boisbaudran), Ho. 16, Er. 8; Yb. shows none in the visible part of the 
spectrum, but many have been mapped in the ultra red portion. 
The qualitative application of this knowledge is as follows :—A mix- 
ture, say of Di. and Ce., is to be freed from Di., this being most familiar 
to myself. The oxalate, first well-washed in dilute HCl, is dried and 
ignited to the condition of an oxide, and then dissolved in HNO, and a 
preliminary examination is made with the spectroscope to make sure of 
the presence of Di., and incidently to judge of its abundance. The 
intensity of that band known as the @ band is especially noted. It lies 
near the Na. band (yellow.) The mixed solution is then subjected to one 
of the separation processes, and as strong a solution in as deep a layer as 
possible, is examined with the spectroscope and the intensity of the @ 
band is noted. This process is repeated until the most intense band has 
entirely disappeared. When I remind you, however, that the removal 
of the Di. by what was undoubtedly the best method known until 
lately, required a seven times repetition the Di. bands growing gradually 
fainter and dying out one by one wntil what had been the most intense 
finally faded entirely from view it will be seen that this fractional 
method forces one to the conclusion that possibly all the Di, has not 
