general promote their moral health,
it is not solely or even chiefly, as lecturers, but as investigators,
that your highest men ought to be employed. You have scientific genius
amongst you--not sown broadcast, believe me, it is sown thus
nowhere--but still scattered here and there. Take all unnecessary
impediments out of its way. Keep your sympathetic eye upon the
originator of knowledge. Give him the freedom necessary for his
researches, not overloading him, either with the duties of tuition or
of administration, nor demanding from him so-called practical
results--above all things, avoiding that question which ignorance so
often addresses to genius: 'What is the use of your work?' Let him
make truth his object, however unpractical for the time being it may
appear. If you cast your bread thus upon the waters, be assured it
will return to you, though it be after many days.

APPENDIX.

ON THE SPECTRA OF POLARIZED LIGHT.

Mr. William Spottiswoode introduced some years ago to the members of
the Royal Institution, in a very striking form, a series of
experiments on the spectra of polarized light. With his large Nicol
prisms he in the first place repeated and explained the experiments of
Foucault and Fizeau, and subsequently enriched the subject by very
beautiful additions of his own. I here append a portion of the
abstract of his discourse:--

'It is well known that if a plate of selenite sufficiently thin be
placed between two Nicol's prisms, or, more technically speaking,
between a polarizer and analyzer, colour will be produced. And the
question proposed is, What is the nature of that colour? is it
simply a pure colour of the spectrum, or is it a compound, and if
so, what are its component parts? The answer given by the wave
theory is in brief this: In its passage through the selenite plate
the rays have been so separated in the direction of their vibrations
and in the velocity of their transmission, that, when re-compounded
by means of the analyzer, they have in some instances neutralized
one another. If this be the case, the fact ought to be visible when
the beam emerging from the analyzer is dispersed by the prism; for
then we have the rays of all the different colours ranged side by
side, and, if any be wanting, their absence will be shown by the
appearance of a dark band in their place in the spectrum. But not
only so; the spectrum ought al