390 
PROF. LOUIS VESSOT KINO ON THE CONVECTION OF 
a, bright red when the fine platinum wires could be easily fused in their proper places ; 
this preliminary heating also served to anneal the wire before its resistance was 
determined. A small weight, attached by means of a silk thread to the lower current 
terminal C 2 , served to keep the wire under slight tension while the distance between 
the potential terminals was accurately measured. 
The whirling table was operated through suitable reducing gear by means of a 
\ h.p., 1200 r.p.m., 110-volt D.C. motor, with speed varied by means of an adjustable 
resistance in series with the armature, controlled by a rheostat conveniently situated 
near the bridge. The field of the motor was separately excited : in order to obtain 
constant speed it was found necessary to excite the field some time before commencing 
a series of observations in order that a steady temperature and therefore steady 
resistance in the field-coils shall have been attained. Another advantage is the 
elimination of disturbing effects on the galvanometer as the motor is stopped and 
started. A timing device attached to the driving shaft of the whirling table caused 
a contact to be made every tenth revolution of the rotating arm and to be registered 
on a chronograph simultaneously with the beats of a seconds pendulum. An examina 
tion of the chronograph sheet showed that with the above precautions the speed 
remained constant to about one-tenth of 1 per cent., well within the range of accuracy 
of the present observations. 
Section 10. 
(i.) On the Constants oj the Platinum. Wires under Test. 
The most difficult measurement in the present study of heat convection and that 
which imposes a limit to the accuracy attainable is the calculation of the temperature 
of the wire from the change of resistance. In fact this point confines the present 
experiment to the use of platinum wires whose constants in the measurement of high 
temperatures are well known. It must be kept in mind, however, that the constants 
of a platinum wire which has been kept for some time at a high temperature for 
a considerable time as in the present experiments are liable to change. This source of 
uncertainty is aggravated in the case of very fine wires where the additional difficulty 
of a sensible change of resistance due to “ evaporation ” from the wire is to be met 
with.( 28 ) Evidence on both these points will be noticed from an inspection of Table II., 
and the limit of accuracy due to these and other difficulties may be roughly set at 
^ of 1 per cent. 
The wires employed in the present experiment were drawn through diamond dies( 29 ) 
from a length of 6 mil pure platinum wire whose constants when used in platinum 
( 2S ) See Burgess and Le Chatelier, ‘ The Measurement; of High Temperatures ’ (Wiley and Sons, 
New York, 1912), p. 232 ; also p. 196. 
(- 9 ) The writer is indebted to the Imperial Wire and Cable Co., of Montreal, for the use of diamond 
dies employed in drawing down the wires. 
