EXPERIMENTS OiN THE CONDENSATION OF STEAM. 



421 



condensation of steam, the temperature of the inner surface of the metal 

 was calculated on the assumption of a rate of condensation equivalent to 

 0'74 T.U.F. per second per square foot per degree F. difference of 

 temperature. The values so found agreed with the observed temperatures 

 within the limits of error of the observations. Owing to the inferior 

 conductivity of the iron, the test was not altogether satisfactory, as the 

 difference of temperature between the steam and the surface rarely 

 amounted to as much as 30 degrees. With a cylinder of pure copper, and 

 using thermo-couples for determining the temperature at a given depth, it 

 should be possible to obtain a more certain confirmation by this method. 



In performing the experiments, a number of variations in points of 

 detail were introduced from time to time. The flow of the circulating 

 water was varied in velocity, and directed in different ways. In order to 

 secure uniformity in the distribution of temperature measured in different 

 directions from the centre, the spiral circulation was found to be essential. 

 In the second apparatus, the screw thread was at first replaced by a baffle 

 plate, which was intended to direct the water into a spiral course, but the 

 results found were unsatisfactory. 



In some cases steam was admitted from the top of the apparatus, and 

 in other cases from the bottom. With the steam supply at the bottom, it 

 was found that the condensed water refused to drain down the vertical 

 1 inch tube in opposition to the current of steam, although the maximum 

 velocity of the steam could not have exceeded 10 feet per second. 



The following set of observations, each of which represents the mean 

 of several taken on different days under similar conditions, will sufficiently 

 indicate the general nature of the results. 



Condensation Results Summary. Mild Steel Bar. Wijyer Removed. 



The temperatures of the metal at distances of 1 inch, 1'5 inch, and 

 2 inches, from the axis of the bar, were observed by means of mercury 

 thermometers which were very cai'efully centred by small iron washers 

 in holes filled with mercury. The hole fitting the bulb of the 1 inch 

 thermometer was j^.- inch in diameter. The other holes were -i\ inch. 



It will be observed that in this particular set of experiments, the 

 temperatures at 1 inch in the metal, when calculated to agi-ee with the 

 assumed rate of condensation, are all too low as compared with those 

 observed, whereas the temperatures similarly calculated at 1 '5 inch are all 

 too high. This might at first sight appear to indicate a very rapid 

 diminution of the conductivity with rise of temperature, but, after making 

 various tests, the effect was traced partly to the disturbance of the heat 

 flow caused by the presence of the holes, and partly to difierences of 

 density of the bar in directions at right angles. The latter differences 

 were not observable in the case of the cast iron. 



