is practically the best.  It should have plenty of depth, so that the card may traverse freely, even when the instrument is inclined:  it should be light in weight, that it may not be easily jarred by a blow; the catch that relieves the card, when the instrument is closed, should be self-acting and should act well:  lastly the movements of the needle should be quick; one that makes slow oscillations should be peremptorily refused, whatever its other merits may be:  the graduation of the degrees on the card should be from 0 degrees to 360 degrees, North being 0 degrees and East 90 degrees.  I wish some optician would make aluminum cards.  The material can be procured as foil, like tinfoil.  It can then be stamped and embossed, in which case it retains its shape perfectly, but I cannot satisfy myself as to a good pattern, nor do I see how to make the North and South halves of the disc sufficiently different in appearance.

Compass for use at night.—­The great majority of compasses are well-nigh useless in the dark, that is, when it is most important to be able to consult them.  They are rarely so constructed, that the difference between the north and south sides is visible by moonlight or by the light of a cigar or piece of tinder.  The more modern contrivances are very effective; in these the southern half of the compass card is painted black, the northern being left white.  With a very faint light, this difference can be appreciated.  In compasses consisting simply of a needle, the north end of the needle should have a conspicuous arrow-head.  It is extraordinary how much the power of seeing a compass or a watch at night is increased by looking nearly at it through a magnifying-glass.  Thus, young people who can focus their vision through a wide range may be observed poring with their eyes close to their books when the light wanes.  So again, at night-time, a placard, even in large type, is illegible at a short distance, but easily read on approaching it.  It seems, in order that a faint image on the retina should be appreciated by the nerves of sight, that image must have considerable extent.

Moonlight or the light of a cigar may be condensed on the compass by a burning glass, or other substitute for it. (See “Burning Glass.”)

True and Magnetic Bearings.—­The confusion between true and magnetic bearings is a continual trouble, even to the most experienced travellers.  Sir Thomas Mitchell’s exploring party very nearly sustained a loss by mistaking the one for the other.  I recommend that the points of the compass, viz.  North, N.N.E., etc., should be solely used for the traveller for his true bearings; and the degrees, as 25 degrees (or N. 25 degrees E.), for his magnetic.  There would then be no reason why the two nomenclatures should interfere with one another, for a traveller’s recollection of the lay of a country depends entirely upon true bearings—­or sunrise, sunset, and the stars—­and is expressed by North,