1914-] 



ATMOSPHERIC PRESSURE, 



69 



would be a marked difference. It will be seen in Fig. 7 that the 

 small circles on the 2.04 megabar line, representing the dried air test, 

 fairly coincide with the small crosses representing the air test in the 

 presence of water. The same is true for the 2.02 megabar line of 

 Fig. 8. Consequently, the effect of moisture in the heat convection 

 of moving air has not yet been determined from our tests, although 

 it would seem reasonable that in view of the very appreciable known 

 thermal capactiy of aqueous vapor, the effect of moisture might have 

 been apparent. 



.^ 



Fig. 9. Graphs of ( — ] against v at dififerent pressures for (9 = 390° C. 



logarithm paper. 



It will be seen that the observations in Figs. 7 and 8 indicate a 

 relation : 



f ^cY ,2/ X / ab watts per cm. \- ^ ^ 



where v^ is a velocity in the neighborhood of 30 cm. per second, which 

 may be assumed as the empirical correction due to free convection 



