280 Transactions of the Boyal Mieroseopieal Society. 
seen when the solution is acid, or alkaline ; as, for instance, in the 
spectrum of the colouring matter of alkanet root, shown by Mr. 
Sorby in his interesting paper on " New and Improved Microscope 
Spf^ctrum Apparatus, and on its Application to various Branches of 
Kesearch."* Also another somewhat remarkable change occurs 
when, by reason of the light reflected through it, the object under 
observation becomes warm, those rays which are the most refran- 
gible nearly always gain in intensity. Nitric totroxide is a good 
example of this. 
With regard to those bands which are situated in the red and 
violet extremities, when measuring them it is as well to direct the 
line of vision so that one sees on the right of such a band equally 
as many divisions of the micrometer as on the left, taking care, at 
the same time, that the eye is placed as near to the eye-piece 
as possible, and kept in one position, or the original position of the 
band may appear to sustain an alteration. Again, as we have pre- 
viously seen in Kule 3, when the slit is too open, or the source 
of illumination too strong, the bands become distorted ; for the 
purpose, therefore, of testing the accuracy of your measurement, it 
is as well to make an observation on some other band, which is 
known to be situated in the vicinity of the one you were work- 
ing on. 
I have likewise been induced, through the ardent researches of 
our President, to reduce my measurements into wave-lengths, and 
having succeeded in working out the divisions of my micrometer to 
that scale, I have, as will be seen, expressed them on my drawings 
by the Greek letter X, w^hilst the micrometer is notified by the 
letter M. By this method I find that the wave-length corre- 
sponding to the position determined by the micrometer, as the 
centre of a large, well-defined band, is not exactly the mean wave- 
length of the two extremities. For instance, note the increased 
size of the wave-lengths in the violet end of the spectrum to those 
in ihe red: the left half of the band seen in the spectroscope 
shows, therefore, more divisions of wave-lengths than that half to 
the right. 
In the description of the micrometer-scale divisions I have 
employed a decimal, but when at work it is more convenient to use 
a fraction of some sort, such as for instance, ^, yq, &c. 
Abbreviations and signs, instead of words, which would take too 
much time to write in full, are very useful whilst making an 
observation ; thus, to denote all well-marked bands, or those of the 
first class, viz, symmetrical, I make use of an M ; whilst for those 
of the second class, or unsymmetrical ones, I use an x. Then, 
again, by reason of the milled head, I am enabled to measure with 
it and the micrometer scale combined to the ten-thousandth of a 
* ' Royal Microscopical Journal,' vol, xiii. p. 198. 
