208 L. HARGRAVE. 
A few experiments have been made at Stanwell Park to show 
how this affects the effective direction of the wind when soaring, 
with the result as I previously surmised, that it was found to 
create a vortex, and that the direction of the air current beneath 
the wing was that indicated by the arrows in (Plate 17, fig. 1). 
The apparatus used was a small bellows, a bent piece of sheet 
aluminium and a candle. The centre of the vortex was found to 
be approximately at the centre of the curve of the fore part of 
the aluminium sheet (Plate 17, fig. 2). The candle you observe 
is not masked by the leading edge. 
The quasi-wing was then bent like (Plate 17, fig. 3), and the 
candle flame was blown in a manner showing that in this case the 
vortex was elliptic. The pressure at A must be greater than at 
C or the candle flame would blow parallel to the blast. Asa first 
attempt to show that the pressure at A was greater than at B, I 
cut a small hole at B and gummed a tissue paper valve opening 
towards B. I could not be certain that air was passing through it. 
A portable forge was now arranged to deliver air through 4 
two inch horizontal tin tube, and various devices were used for 
hanging things in front of the blast. Among others an old and 
rough gull’s wing showed the loose feathers blowing towards the 
front edge (Plate 17, fig. 4). 
When a piece of the wing was cut off and hung at eleven inches 
from the blast with a negative angle of about 28°, it at once began 
to revolve in an elliptic orbit, the feathers on the under side of 
the wing being ruffled back at positions A and B (Plate 17, fig. 5). 
When attempting to repeat this experiment the wing could only 
be made to revolve in a contrary orbit to that shown in the figure- 
A piece of aluminium without a bulb would only swing very 
slightly in the line of blast. 
A piece of tin folded with a bulb at the forward edge and some 
pieces of down gummed on the concave side, was set at a positive 
angle of about 6°, the down at the bulb blew strongly in the direc- 
tion of the arrow (Plate 17, fig. 6). 
