•84 
POPULAR SCIENCE REVIEW. 
to its salts, in order to ascertain if it was erbium itself. Employing the 
residues from which Nilson had separated ytterbia and scandia, he found 
it impossible to obtain a red oxide with a constant molecular weight, 
even after several hundred decompositions. Suspecting the presence of the 
unknown oxide he induced Thalen to examine the spectrum of what he 
regarded as the purest erbia, and to compare it with that of yttria 
and ytterbia. Certain absorption bands in the last fractions suggested that 
the colour of erbia is due to the presence of three oxides giving absorption 
spectra. The reddest of the fractions (RO molecular weight = 125-127) 
were united and submitted to a long series of decompositions, one fraction 
being treated for ytterbia, another for yttria, and a third, intermediate, con- 
taining the concentrated erbia. At the same time, he attempted to concen- 
trate the colouring matter in residues A, rich in ytterbia, and B in yttria. 
After pushing the treatment as far as possible with the amount of material 
at his disposal, he handed over the five fractions to Thalen, who found bands 
common to all the fractions, and due probably to erbia. The following 
bands varied markedly from one fraction to another : — 
Fraction A. 
Erbium ? 
Fraction B. 
Extr. from 
Extr. from 
Mean 
Extr. from 
Extr. from 
Wa/ve 
ytterbia 
Erbia, 
fractions, 
Erbia, 
residues rich 
length. 
residue. 
126-127. 
126-127. 
126-127. 
in yttrium. 
* 6480 
strong 
quite strong 
fails 
fails 
fails 
y 6400-6425 
fails or trace 
trace 
weak 
weak 
pretty strong 
z 5360 
fails 
fails or trace 
trace 
feeble 
quite strong 
Hence x belongs to fractions situated near ytterbia, and does not exist in 
fractions from yttrium. But, on the other hand, y and z fail in the ytterbium 
residues, but grow more distinct as yttrium is approached. The ytterbia 
fractions gave a rose-colour with a tinge of violet ; the yttria fractions had 
an orange tint. For the element coming between ytterbia and erbia, 
characterized by band x in the red of the spectrum, Cleve proposes the name 
of Thulium , from Thule, the ancient name of Scandinavia. The atomic 
weight, Tm, should be about 113 if its oxide be RO.. Erbium proper, which 
has the common bands mentioned, has the atomic weight 110-111. Its 
oxide is of a light rose colour. The third metal, characterized by the y and 
z bands, and which is between erbia and terbia, should have an atomic weight 
below 108. Its oxide appears to be yellow. For it the name of Holmium is 
proposed, derived from the Latin name of Stockholm, the environs of which 
are rich in yttria minerals. — (' Compt . Rend. 1879, lxxxix. 478.) 
Notes on the two new Elements announoed by Cleve . — Soret points out that 
he showed in the spring of 1878 that the two bands which characterize 
holmium do not belong to erbia, but to a new earth which he called 
provisionally X, and which is perhaps identical with philippium, since dis- 
covered by Delafontaine. Besides these two bands, Soret recognized three 
other absorption bands ; one less refrangible than A, a second overlapping 
the band of erbia in the indigo, and a third (faint) in the violet a little beyond 
h. In the ultra violet spectrum six absorption maxima exist between H 
and R. In samarskite the earth X is relatively to erbia much more abundant 
than in gadolinite. As to the red ray which characterizes thulium, Soret 
had alreadv observed that also in some ytterbia products which had been 
