286 
modation for the work will be provided in a 
new three-story building to be erected along- 
side the Administration Building, but mean- 
time a house close to the laboratory entrance 
is being equipped temporarily for the purpose. 
A watch sent for testing by Paul Ditisheim, 
La Chaux de Fonds, Switzerland, obtained 96.2 
marks out of a possible 100—the highest total 
yet recorded in the Kew trials. 
The investigations of the methods of notched 
bar impact testing were continued, and in the 
present year it is proposed to inquire into the 
following questions: Need for test bars of dif- 
ferent sizes; the question of test bars supported 
at two ends or fixed at one end; how far the 
work of fracture depends on the cross-section 
or the volume of the test bar, or both, and 
whether a relation can be found for reducing 
results on one size of bar to those on another; 
influence of the velocity of striking and angle 
of bend of the bar; and application of notched 
bar tests to carbon steels in industrial prod- 
ucts. The hardness test research is being ex- 
tended to cover the effect of the wear caused 
by sliding abrasion, of variations in the rela- 
tive velocity of the surfaces, and of the pres- 
sure between them. ‘he Brinell ball test, 
usually adopted for indicating the indenta- 
tion hardness of a material, is open to the ob- 
jection that it varies with the load and with 
the hardness of ball used. Hence a standard 
ball of 10 mm. diameter and a fixed load of 
3,000 kg. are employed, but it is not always 
possible to use this standard—e. g., with very 
thin test pieces or very soft materials. Ex- 
periments have therefore been made with the 
object of studying the various methods of 
bringing results obtained with different balls 
and loads into line with those obtained by 
using the standard. 
The experiments with the apparatus described 
in last year’s report for determining the heat 
transmission to water from the internal sur- 
faces of brass pipes, smooth and roughened, 
were completed. The results showed that, 
with the amount of roughening obtained, the 
heat transmission per unit surface per degree 
in difference of temperature between metal 
and water for the smooth pipe could be in- 
SCIENCE 
[N. S. Vou. XLVITI. No. 1238 
creased in the ratio of about 2.5 to 1 for the 
same mean yelocity of flow. It was obvious 
that if improvements of this order could be 
made in the surfaces of air-cooled engines and 
radiators, considerable economy of material 
could be effected in aeroplane engine design. 
To test the question a series of copper gills of 
the form and distribution commonly adopted 
in air-cooled engines was fixed to the cylinder 
of the engine and set up in a wind channel. 
An electric current was then circulated 
through the gills, and from measurements of 
the current and resistance it was possible to 
ascertain the temperature of the gills and the 
heat dissipated. The gills were first tested in 
their ordinary smooth condition and then 
roughened by means of corrugated steel dies. 
The roughening, however, appeared to produce 
practically no effect on the heat transmission, 
this result indicating that the convection of 
momentum to the rough and smooth surfaces 
was approximately the same. The reason for 
this apparent discrepancy between the two sets 
of observations is still under investigation. 
The demands of the Aerodynamics Depart- 
ment are so numerous and important that the 
erection of two new air channels has been 
called for. A new airsecreen balance for the 
7 foot tunnel is in hand and a new apparatus 
for lift measurement is in use. Modified ap- 
paratus for the measurement of rotary deriva- 
tives was devised and will be further de- 
veloped. 
The research work carried out was of the 
most varied character. Tests were made on 
models of aeroplane wings, monoplane and bi- 
plane models of complete aeroplanes, airships 
and kite balloon models, and models of air- 
screws, with calculations relating to stability, 
streneth of construction, bomb-dropping, ete. 
A large amount of work was done in con- 
nection with the design of wind channels, and 
the research on eddy motion was continued. 
Many of the investigations were made in 
response to direct requests from the Air Min- 
istry, but every effort was made to increase 
the value of special tests by bringing them into 
proper relation as part of an organized scheme 
of research. 
