As I returned from my walk the southern sky seemed to grow darker, and later stratus cloud was undoubtedly spreading up from that direction—­this at about 5 P.M.  About 7 a moderate north wind sprang up.  This seemed to indicate a southerly blow, and at about 9 the wind shifted to that quarter and blew gustily, 25 to 35 m.p.h.  One cannot see the result on the Strait, but I fear it means that the ice has gone out again in places.  The wind dropped as suddenly as it had arisen soon after midnight.

In the evening Simpson gave us his first meteorological lecture—­the subject, ‘Coronas, Halos, Rainbows, and Auroras.’  He has a remarkable power of exposition and taught me more of these phenomena in the hour than I had learnt by all previous interested inquiries concerning them.

I note one or two points concerning each phenomenon.

Corona.—­White to brown inside ring called Aureola—­outside are sometimes seen two or three rings of prismatic light in addition.  Caused by diffraction of light round drops of water or ice crystals; diameter of rings inversely proportionate to size of drops or crystals—­mixed sizes of ditto causes aureola without rings.

Halos.—­Caused by refraction and reflection through and from ice crystals.  In this connection the hexagonal, tetrahedonal type of crystallisation is first to be noted; then the infinite number of forms in which this can be modified together with result of fractures:  two forms predominate, the plate and the needle; these forms falling through air assume definite position—­the plate falls horizontally swaying to and fro, the needle turns rapidly about its longer axis, which remains horizontal.  Simpson showed excellent experiments to illustrate; consideration of these facts and refraction of light striking crystals clearly leads to explanation of various complicated halo phenomena such as recorded and such as seen by us on the Great Barrier, and draws attention to the critical refraction angles of 32 deg. and 46 deg., the radius of inner and outer rings, the position of mock suns, contra suns, zenith circles, &c.

Further measurements are needed; for instance of streamers from mock suns and examination of ice crystals. (Record of ice crystals seen on Barrier Surface.)

Rainbows.—­Caused by reflection and refraction from and through drops of water—­colours vary with size of drops, the smaller the drop the lighter the colours and nearer to the violet end of the spectrum—­hence white rainbow as seen on the Barrier, very small drops.

Double Bows—­diameters must be 84 deg. and 100 deg.—­again from laws of refraction—­colours:  inner, red outside; outer, red inside—­i.e. reds come together.

Wanted to see more rainbows on Barrier.  In this connection a good rainbow was seen to N.W. in February from winter quarters.  Reports should note colours and relative width of bands of colour.

Iridescent Clouds.—­Not yet understood; observations required, especially angular distance from the sun.