1919.] New Zealand Institute Science Congress. 301 
The existence of cloud in the mass of air crossing the divide mast 
prevent some lowering of temperature by radiation, and again after the 
ranges are left behind and the air becomes once more transparent its higher 
portions gradually cool by radiation, ultimately reaching the dew-point, 
and a cloud stratum is again formed over the plain, with a well-defined 
edge towards the hills, and we see the familiar north-west arch from any 
spot nearer the sea. 
These must be the general phenomena ; but conditions over the Tasman 
Sea are variable, and it is probable that sometimes the north-west wind 
off that sea is comparatively dry, and the phenomena might be very much 
less in degree, so much so that a somewhat cold north-west wind may be 
experienced on the eastern side. 
The correctness or otherwise of the explanation might be verified 
by the temporary establishment of a chain of meteorological posts of 
observation across the island, to give sets of simultaneous observations. 
It is not, of course, to be assumed that a north-west wind here does 
always come straight across the Island. The effect will be the same, 
though, provided the condition attained in the height of its rise is that 
outlined above. 
Undoubtedly some criticism of the above explanation will be forth¬ 
coming, and in view of the great popularity of the weather as a subject 
for discussion one need not be surprised if some one blames Maxwell’s 
demons for the dust. 
The Harmonic Analysis of Tidal Observations and the Prediction of 
Tides, by C. E. Adams. 
Abstract. 
Practically all existing methods used for the harmonic analysis of tidal 
observations have many numerical approximations employed in their prac¬ 
tical applications. It therefore becomes difficult to determine whether the 
method of harmonic analysis gives the best representation of the tide 
curve. 
In the method here described no arithmetical approximations are used, 
so that a criterion is obtained to test the application of Fourier’s series to 
tidal observations. 
A description is given of the method of prediction of tides used in New 
Zealand. An illustrated description of this process appeared in the Survey 
Reports of the Lands and Survey Department for the years 1910-14, and 
to these reports reference should be made for the details. The process is 
a graphic one, controlled by calculation. The results of the predictions 
for Wellington and Auckland are published in the New Zealand Nautical 
AImanac by the Marine Department. 
A Slide-rule for solving the Quadratic Equation, by D. M. Y. Sommer- 
VILLE. 
Press Notice (Lyttelton Times ). 
Professor D. M. Y. Sommerville, Wellington, exhibited and described! a slide-rule 
specially constructed for finding the roots of a quadratic equation. In principle it con¬ 
sisted of two movable scales sliding side by side between two fixed s°a< . The movable 
scales are set to the values of the two terms of the equation, and the roots of the 
equation are obtained by reading off the coincident numbers on the two slides. 
