THE RARE EARTHS.—MAGEE. — Ixx1 
been removed, but that a more powerful spectroscope in the hands of 
very experienced men might rediscover, in what appeared free from Di, 
traces of it yet present. This has actually happened more than once, and 
was what caused Prof. Dennis and myself to lay so much stress upon our 
new separation process, for the removal of Di. from Ce. We believe 
that the new method removes the Di. at once and entirely ; for with the 
best spectroscope at our disposal we could find no trace of Di. The pro- 
cess in other words is not fractional but immediate. I can perhaps 
illustrate this better by an example, familiar perhaps to all of us. Ferro- 
and Ferri-Cyanides of K. long served as accurate tests of Fe., later Pot- 
assium Sulphocyanate was found to detect ferric Fe, when the Cynanides 
failed todo so. This reagent gives to a solution containing ferric Fe. in 
solution, a blood red or, in weaker solution, a wine color. When, how- 
ever, some chemist proposed to add ether to the solution after testing for 
Fe. and failing to obtain a color, he found on closing the test-tube, and 
shaking violently that from a solution that was colorless after adding 
KCNS, a red color was extracted by the ether. This of course gives a 
vary dolicate test for Fe., a delicacy unsought for a few years since. 
The Di. test is probably not so delicate at present. 
But just here comes in one more of the evil features of the rare earth 
work, for the test that serves to prove the absence*of Di must serve also 
as the test for La. which has no absorption bands, since La. being more 
strongly basic than Di., when the latter is known to be removed the 
former must have been previously gotten rid of. Of course the spark 
spectra could be employed, and, unless some easier method is discovered, 
must be employed in very accurate work, but it is tedious and requires 
special apparatus and precludes all workers, but those who have the 
advantages of the finest university laboratories, or are themselves 
wealthy. It needs, moreover, a much longer training than is needed to 
use the absorption band method. It is seldom employed as a test. The 
other method, applicable also in every case, but slow and requiring the 
very highest chemical skill to ensure results is to make equivalent weight 
determinations. This has so far been done gravimetrically, but methods 
are being sought by which it may be done volumetrically, which will be 
a great shortening, and in skilled hands, if the methods are good, will 
yield excellent results. 
Having now given a fair idea of what the rare earths are, how they 
act chemically, and the difficulty of experimental work with them, I will 
