4G PROFESSOR A. CRICHTON MITCHELL ON THE 



the stream lines of fumes in passing round the ball. The results were, however, indecisive. 

 Perhaps a better indication as to whether such a critical speed really existed is afforded 

 in another way. I have already stated that the speed recorded by the anemometer 

 when the ball was hanging in the tube was less than when the ball was not in the tube, 

 and that this difference varied with the speed. If now a curve be drawn whose ordinates 

 represent this difference, and whose abscissae represent the speeds with the ball in the 

 tube, it will be seen that when a speed of about 450 metres per minute is reached the 

 ordinates begin to increase much more quickly. This may be due to the motion of the 

 air in passing, and after having passed, the ball ceasing to be steady at this speed, and 

 therefore recording less in the anemometer. If, now, the curves in fig. 6 be examined, 

 it will be seen that up to about the same speed the rate of cooling is nearly proportional 

 to speed, but beyond that it is less than proportional to speed. I am therefore inclined 

 to think that the change in curvature in the curves of fig. 6 is due to the motion of the 

 air ceasing to be steady at or about a speed of 450 metres per minute. 



Of course it is to be remembered that these results apply only to a cooling body of 

 the shape and size of that used in this investigation. Were either shape or size 

 different, the results would be different. 



An improved method of experiment would be to heat a strip of platinum foil by 

 means of an electric current, allow it to cool while exposed to a current of air passing 

 across its breadth, and ascertain its temperature from time to time by means of its 

 electrical resistance. It would then be possible to determine in absolute measure the 

 amount of heat lost in unit time from unit surface for different excesses of temperature 

 and for different speeds of air. No question regarding the character of the motion, 

 steady or otherwise, of the air would then be involved. 



There is another aspect of the question which deserves consideration. If the speed 

 of the air current were increased enormously, friction between the air and the ball 

 would generate heat, and thereby lessen the rate of cooling. That such would be the 

 case is known from the behaviour of meteors in passing through the earth's atmosphere. 

 Now, if the friction between highly rarefied air and a meteoric body moving at, say, 

 20 miles per second, is sufficient to render the body incandescent, the amount of heat 

 similarly generated in air at the earth's surface with a speed of 45 miles per hour (a 

 speed attained easily by the apparatus employed) may be sufficiently large to be 

 measurable by experimental means. 1 tried to detect any result of this kind by the 

 following experiment. The ball was placed in the tube, and allowed to remain there 

 for forty-eight hours, so that its temperature might become the same as that of its 

 surroundings and of the air in the room.* A Boys' radio -micrometer was fitted up in 

 such a position that any rise in temperature of the surface of the ball might be at once 

 detected. A current of air with a speed of about 45 miles per hour was then allowed 

 to pass along the tube for ten minutes, after which the screen between the ball and 



* The experiments described in this paper were conducted in an underground cellar, in which the diurnal variation 

 of temperature is scarcely noticeable. 



