446 Prof. L. Hill and Mr. D. Hargood Ash. 



where SH is the increment of heat loss due to a small change in pressure 

 8p, that is to say, the percentage change in rate of heat loss is one half the 

 percentage change in pressure. 



We have further investigated this matter in chambers constructed for high 

 pressure and for low pressure observations, into which we could go ourselves 

 together with our apparatus, chambers kindly put at our use by Mr. E. H. 

 Davis, of Messrs. Siebe, Gorman, Ltd., to whom we are much indebted for the 

 help thus rendered. 



Since the theory of the loss of heat by the kata by convection shows that 

 the rate of heat loss is proportional to the square root of the density, and 

 therefore of the pressure, other conditions remaining constant, we may write 



H, = Cyp, (i) 

 where = heat lost by convection, p = pressure, and C = a constant. 

 Experimental evidence shows that at ordinary temperatures and pressures 

 half the heat loss of the kata is due to convection and half to 'radiation 

 Assuming this to be case if H is the total heat lost, He that lost by convection, 

 and that lost by radiation, we have, since H = 0"27 (as proved in the 

 Eoyal Society paper already cited), 



H« = (u) = (ill) 



also from (i) 



substituting for H^j in (iv) from (ii), 



where Hi is the total heat lost at a pressure pi. 



Now the loss of heat by radiation will be unaltered by changes of pressure, 

 therefore if Hg is the total heat lost at a pressure P2, H(.j being the loss due 

 to convection, then 



H2 = H,, + H,,, (vii) 

 where H, is the heat lost by radiation at a pressure pi ; hence substituting 

 in (vii) for "K^ from (vi), 



2 V y,' 



