20tj Transactions of the Royal Microscopical Society. 
chemically modified. Haemoglobin is a very instructive example. 
As is so well known, the oxidized modification gives a spectrum with 
two well-marked absorption bands, and on passing through the 
solution a stream of carbonic oxide, the loosely combined oxygen is 
replaced, and the spectrum shows two absorption bands, very 
similar to the original, rather nearer to the blue end. On the 
contrary, when the oxygen is replaced by passing nitric oxide 
through an ammoniacal solution, though we still see two such bands, 
they lie nearer to the red end. In all three cases, however, the 
ratio between the wave-lengths of the centre of the two bands is, 
very nearly, if not absolutely, the same. The changes, therefore, 
are analogous to what would occur if we were to strike two strings 
on a harp, and then strike them again, raising or lowering their 
jfitch by means of the pedals, the effect of the substitution of the 
oxygen by carbonic oxide being as it were to make the bands more 
sharp, whilst nitric oxide makes them more flat. The removal of 
the oxygen produces a total change, and is thus unlike mere 
substitution. 
6. Probable Evidence of Chemical Relationships afforded by 
Spectra. 
Such, then, being the relation between the spectra of compounds 
which are known to be related in a very simple manner, and can be 
changed one into the other, it becomes a question of much interest 
to consider whether when we meet with spectra having similar 
relations the substances may not be in some way connected, 
although it may be impossible to convert one into the other. 
I find that by different processes two different modifications of 
liaematin may be prepared, which give very similar spectra, but 
the absorption bands are in very different positions, though at the 
same wave-length ratio. On deoxidizing both by means of a little 
of the sulphate of protoxide of iron and ammonia, both are altered 
and give spectra similarly related, but by deoxidizing with sulphide 
of ammonium both are changed into the same substance and give 
the same spectrum. Here, then, we see that these two kinds of 
haematin must be connected in some simple manner, and by a 
strong deoxidizing reagent can be both changed into the same sub- 
stance. There are, however, cases in which two substances give 
spectra having the above-named characters, and on adding various 
reagents both are altered in the same manner, but yet, so far, it is 
not known that one can he changed into the other, or both into 
some common product. Various plants furnish illustrations of this 
fact, but I do not know any more decided than the purple colouring 
matter of Lobelia speciosa as compared with the pink colour of 
