72 
POPULAR -SCIENCE REVIEW. 
compare the spectrum of different thicknesses of a known 
material, with that of blow-pipe beads coloured with some 
unknown substance. 
Though perhaps not in all cases, yet very often, a reduction 
in the amount of light transmitted through the slit is equiva- 
lent to an increase in the thickness of the object. Therefore, 
by varying the width of the opening in the stage attached to 
the eye-piece, we modify the spectrum much as if we altered 
the thickness of the object. To some extent, the same effect 
is produced by altering the width of the slit inside the eye- 
piece, but then the spectrum is also modified in other 
respects, and well-defined absorption bands are made less so 
if the slit be too wide. At the same time, some wide and 
obscure bands are far more distinct when the slit is wide ; a 
wider slit and greater dispersion being in some degree equiva- 
lent to a narrower slit and less dispersion. 
Having now, I trust, given a sufficient general account of 
the preparation of objects, I will proceed to describe a few 
particular applications of the instrument; and since I have just 
referred to blow-pipe beads, — and that is a new branch of the 
subject, — it will perhaps be well to say a few words on it now. 
I shall only give a few illustrations, because I have, so far, 
done little more than convince myself that in some cases the 
instrument may be very advantageously employed. The blue 
modifications of salts of cobalt present us with very character- 
istic spectra. The pink hydrated crystals of the chloride give 
two well-marked absorption bands, one in the light part, 
the green, and the other in the blue green, as shown in 
fig. 4, b; but when dissolved in water there is only one 
broad absorption band in the centre of the green, as shown in 
a , which cannot be looked upon as characteristic. If dis- 
solved in a concentrated solution of chloride of sodium, the 
position of this absorption band is not materially changed, 
though the general colour is more red; but if dissolved 
in a concentrated solution of chloride of calcium, there is 
a complete change. We obtain a blue solution, which gives 
a most characteristic spectrum when there is only suffi- 
cient cobalt present to give a very pale blue tint. To see it 
to advantage, lamp-light should be used. We then get a 
spectrum somewhat as shown in fig. 4, c, which is, however, 
not exactly like what is seen with any one case, but shows 
the characteristic peculiarities of different intensities of colour. 
There is an absorption band in the red which may be sepa- 
rated into two ; another in the lower part of the red, and 
another more faint in the orange. These two bands are so 
very sharp and narrow, that they look almost like the so-called 
Fraunhofer* s lines seen in the spectrum of day-light. To see 
