369 
or less strong yellow, when either of the above-named sub- 
stances is present. The etherial solution should he trans- 
ferred to a small evaporating dish by means of a pipette, and 
a fresh quantity of ether agitated with the wine, and added 
to the other. The most useful kind of pipette for this and 
similar purposes is one with a small vulcanized india-rubber 
top, so arranged that, after having been compressed, it may 
by its own elastic force draw xip the liquid. After evaporat- 
ing the solution to dryness, a small quantity of the colour 
should be dissolved in water in an experiment-cell, and 
treated with bicarbonate of ammonia. In both cases this 
develops a single very distinct absorption-band in the green, 
that characteristic of logwood being situated at of my 
scale, whilst that of Brazil wood is farther from the red end, 
at 54 ; and the solution is strongly fluorescent, of a peculiar 
orange colour. These spectra are so characteristic, and can 
be so easily compared with those of the substances them- 
selves, that an extremely minute quantity of either could be 
detected with certainty. 
The colour of rhatany root does not give any well-marked 
band in an aqueous solution, either acid or alkaline ; but 
when dissolved in alcohol and slightly acid, it shows a 
moderately distinct band between the yellow and green, at 3f 
of my scale, and a fainter at 74. In order to detect this sub- 
stance the Avine should, therefore, be evaporated to small 
bulk, and redissolved in strong alcohol, and after this has 
stood in a test-tube until the insoluble matter has been 
deposited and the solution has become quite clear, the 
absorption-band at 3f may be more or less distinctly recog- 
nised, according to the amount of the rhatany root present ; 
but the natural colour of the wine makes it impossible to 
detect a very small quantity. 
The dark crimson berries of the Virginian poke are remark- 
able for containing a colour belonging to my group C ; 
whereas those of nearly all fruits of that tint belong to 
group B. It shoAvs two absorption-bands, Avhich are more 
distinct in an alcoholic than in an aqueous solution, and are 
4f, 7f. To detect this substance Ave ought, therefore, to 
adopt the same process as in the case of rhatany root, and 
endeavour to see those bands, Avhich might sometimes be 
made more distinct by adding a little Avater and sulphite 
of soda. 
These substances may be changed by keeping long in 
solution, and therefore might not be detected in old Avines. 
