54 The Philippine Journal of Science im 



Although it has been fairly well established that the humidity 

 of the air does not affect the efficiency of the charcoal as an 

 absorber of the emanation, still we deemed it advisable to dry 

 the air thoroughly before it reached the charcoal. The air 

 during the greater part of the year in Manila has an extremely 

 high absolute humidity, and the presence of a great amount of 

 moisture in the charcoal is very annoying when it comes to 

 heating the tubes. Since neither sulphuric acid nor calcium 

 chloride seem to absorb the emanation, we first passed the air 

 through a bottle containing sulphuric acid and then through 

 tubes filled with calcium chloride. Most of the water was ex- 

 tracted from the air by the acid which was renewed about once 

 a month. 



The charcoal tubes. — The first tubes tried were of fused quartz, 



1 meter long, having diameters of about 1.2 and 1.6 centimeters, 

 respectively. The inequality in the diameters made it very dif- 

 ficult to regulate the amount of charcoal in the tubes so that 

 the absorption in the two branches was exactly equal. Conse- 

 quently, after a few preliminary experiments, these tubes were 

 replaced by electrosilica tubes, 60 centimeters long and having 

 a uniform bore of 1.5 centimeters. These tubes were filled to 

 within about 6 centimeters of the ends with granulated coconut 

 charcoal, each tube holding 70 grams. Two of these tubes were 

 placed in series in each branch of the experiment, so that the 

 air had to pass through 140 grams of charcoal. After collecting 

 the emanation, the tubes which had been connected in series 

 were heated in parallel in a tubular electric furnace to a bright 

 red heat. Tests were made on the relative absorption of the 

 two sets of tubes, no difference being detected. 



The manometers. — The rate of flow of the air through the 

 tubes was measured by means of water manometers across glass 

 capillary tubes. The manometers were carefully graduated for 

 different rates of flow, and the corresponding curve plotted. 

 The combined error of graduation and reading was not over 



2 per cent. 



The suction pump. — At first a filter pump attached to the city 

 water system was tried, but the water pressure was subject to 

 frequent and rather large variations, and we were compelled 

 to substitute for the filter pump a motor-driven oil pump. By 

 placing in the system a large equilibrating bottle and a mercury 

 regulator, we were able to obtain an almost absolutely steady 

 flow of air for any desired length of time. The rate of flow 

 was regulated by means of pinch cocks placed on the rubber 

 tubing between the pump and manometers. 



