320 FORCING GARIEN. 
or by adding it if to the south.* These periods are of 
course selected in accordance with the time at which the 
direct rays are most required. Mr. Knight proposes a 
general elevation of 34° for the latitude of London, an 
angle which corresponds to the 20th of May and 2ist of 
July. This would afford four months, from the 20th of April 
to the 2ist of August, during which the angle of incidence 
at mid-day would not at any time amount to 9°, while the 
deviation at the winter solstice would be 48°, and the loss 
of light from reflexion would be little more than 1-30. The 
Rev. Mr. Wilkinson recommends: 45°, a pitch extremely 
suitable for early vineries and pine-stoves. In this case, 
the midsummer deviation would be 19°, and the loss 1-40, 
and the midwinter deviation 30°, while the loss is nearly 
the same. From these statements, however, and from an 
inspection of the table already referred to, it is manifest 
that much greater exactness has been sought in this matter 
than is at all necessary. The reduction of the opacity of 
the roof, arising from the breadth and depth of rafters and 
astragals, is of much greater consequence. Accordingly, 
in some glazed houses, particularly those constructed of 
metallic substances, rafters have been omitted altogether ; 
* The following is part of Bouguer’s Table of Refloxions. Of 1000 incé 
dent rays, when the 
Angle of incidents is 75° 299 rays are reflected. 
70 222 
65 157 
60 112 
50 57 
40 34 
30 at 
20 25 
10 25 
1 25 
