544 NATURE 
given by tests in water of ebonite models of 1 ins 
diameter, is of much interest. The difference between 
the densities of the two media, air and water, is not 
a source of difficulty in such comparison : the relative 
resistances are directly proportional to the densities 
of the media, and allowance for the difference in 
density is thus readily made. According to the law 
of dynamical similarity, referred to in previous reports, 
and clearly enunciated by Lord Rayleigh in the report 
for 1909-10, the quantities on which variation in the 
resistance coefficient may be expected to depend are 
the relative dimensions, the relative velocities, and the 
“kinematical viscosities.’’ The velocities in the two 
sets of experiments, made at the Aircraft Factory and 
the National Physical Laboratory respectively, were 
20 ft. per second and 1°78 ft. per second. The kine- 
matical viscosities of air and water. are in the ratio 
of 13 to 1. Employing the law of dynamical similarity 
the ‘two series of experimental determinations enable a 
provisional estimate to be formed of the effect on the 
coefficient of head resistance of change in velocity, 
and of change in dimensions. Mr. Bairstow, of the 
National Physical Laboratory, has made the calcula- 
tion, and employing the data so obtained, has esti- 
mated the resistance of a full-sized balloon, with 
smooth surface, of 40 ft. diameter and of specified 
form, with fineness ratio of 6: 1, when travelling at 
the rate of 40 miles per hour, to be 320 lb. weight. 
To obtain further information on this important 
question of the variation of the resistance coefficient 
with dimensions, a large wooden model of an airship, 
6 ft. in length and 1 ft. in diameter, has been made 
at the laboratory, and its resistance will be deter- 
mined by towing tests in the William Froude National 
Tank. These experiments are now in progress. A 
further model, 4 in. in diameter and 2 ft. long, is also 
under construction for towing tests in the tank, and 
it is hoped that a comparison of the various experi- 
mental results available may lead to valuable conclu- 
sions as to the relation between the resistance of 
models and of the full-scale machines, and may furnish 
data sufficient to enable the prediction, from observa- 
tions on models, of the absolute magnitudes of the 
forces acting on full-sized airships and aéroplanes to 
be made with more confidence than is at present 
possible. 
Investigation of the Pressure Distribution Round a 
Thin Plate and an Aérofoil_—The object of. these ex- 
periments was to examine closely the character of the 
air flow round a thin plate or an aérofoil, and to 
investigate the way in which the total ‘lift’? and 
“drift ’’—apart from friction—on the whole plate are 
built up fromthe pressures, or ‘‘suctions,’’ at different 
regions of the upper and lower surfaces. 
The detailed results and distribution curves, which 
will be given in the Technical Report, exhibit many 
points of interest, and of importance in aéroplane 
design. Thus for the aérofoil tested there was a 
particular angle at which the upper, convex surface 
gave its maximum contribution towards the total lift, 
and another,. different angle at which the under, 
concave surface gave the maximum effect. It thus 
appeared to be a possibility that by variation of one 
of the surfaces improved efficiency could be obtained. 
The nature of the-pressure distribution on the con- 
vex surface of the aérofoil presents some remarkable 
features. At inclinations commonly occurring in flight 
practice, from 5° to 10°, the negative pressure on the 
convex surface is a miami and reaches a_ very 
high value, at a point immediatelv behind the leading 
edge of the ‘‘plane.’’ The same fact is shown in the 
distribution curves for different aérofoils at an angle 
of 6° given by M. Eiffel, who has also carried out a 
large number of experiments in the plotting of pres- 
sure distribution, to which the National Physical 
NO. 2230, VOL. 89] 
, the convex surface, and from 123 
[JULY 25, 1912 
' Laboratory measurements may be regarded as come 
plementary. 
Another interesting feature of the results obtained 
for the aérofoil is that at an inclination of about 123° 
there is a marked change in the pressure intensity on 
° to 20° the conditions 
of flow appear to be so unsteady that no readings of 
the pressure intensity could be made, the pressure 
varying incessantly and erratically within wide limits. 
This critical region is also indicated, in a léss marked 
manner, by the measurements made on the concave 
surface. 
Effect of Separate Variation of the Upper and Lower 
Surfaces of an Aérofoil.—In continuation of the inves- 
tigation above described, into the pressure dis- 
tribution, the effect has been examined of varying 
one surface only of the aérofoil, the curved under 
surface of the aforementioned aérofoil being replaced 
by a plane. 
The general conclusion arrived at is that, as a first 
approximation, each of the surfaces of an aérofoil 
can be independently designed; the second approxima- 
tion, due to interaction between one surface and the 
other, is sufficiently small to be regarded as of the 
nature of a correction. 
The curves obtained for the lift and drift, and the 
ratio of lift to drift, show clearly the effect of replac- 
ing a cambered under surface by a plane one. Over 
the useful range of inclinations from 7° to 12°, ‘the 
ratio of lift to drift is nearly the same for both aéro- 
foils, but the lift coefficient at 10° decreases from 
048 to o'42. It follows from this that about 14 per 
cent. increase in wing area would be required to 
produce the same lift. 
Effect of Variation of the Spacing of the Two Planes 
in a Biplane.—These experiments were made with 
two facsimiles of a wing form of the Blériot type, 
and the ‘‘gap”’ between the two planes was varied 
from o4 to 16 times the breadth of either plane: 
The results were corrected for the resistance of con- 
nections. They show appreciable loss of lift per unit 
area as compared with the single plane; when the 
‘““gap”’ is equal to the breadth of either plane, the 
loss is 17 per cent. Even with a “gap” equal to 16 
times the breadth, the loss is still as much as 10 per 
cent. The ‘‘drift’’ values for the biplane do not differ 
greatly from those for the single plane; the percentage 
losses in the ratio of lift to drift are thus nearly of 
the same magnitude as those in the lift. 
The advantage that might be gained by employing 
a wider spacing than the usual one, with a gap equal 
to or slightly greater than the breadth of a plane, is, 
of course, to some extent, counterbalanced by the in- 
creased resistance and added weight due to the extra 
length of struts necessary. The best spacing depends 
on the conditions of design, and is different if the 
speed be required to be kept constant from that most 
suitable for a machine having wings of fixed area. 
For flight speeds ranging from 40 to 60 miles an 
hour the best biplane spacing is in the neighbourhood 
of that most commonly adopted, with a ‘‘gap”’ ap- 
proximately equal to the chord. 
Effect of Camber.—The effect of variation of the 
camber of the upper surface, and also of the lower 
surface, has been investigated. As already stated, it 
had been previously shown that, to a first approxima- 
tion, the upper and lower surfaces might. be inde- 
pendently designed. The experiments on the variation 
of camber of the upper surface were made on aérofoils 
having their lower surfaces plane. The amount of 
camber of the upper surface giving a maximum value 
of the ratio of lift to drift was found. to be about 1 in 
20, as compared with Eiffel’s value of 1 in 13'5. 
The experiments on the effect of varying the camber 
of the lower surface were made on an aérofoil in 
t 
Qe 
