178 



water and ice, or ice and salt, depending on ttie temperatures required in 

 tlie tanlv T,. 



Tlie water used in this experiment was first distilled in the usual 

 copper still, then with pctassium permanganate in glass, then twice again 

 in glass. Just before using the water was boiled for fifteen minutes to 

 drive off absorbed gases, and then rapidly cooled by placing the flask in 

 ice water. The water was siphoned from the flask through a glass siphon 

 with a cock which permitted the flow to be adjusted at will. Before open- 

 ing the cock the water in the flask was each time brought to approxi- 

 mately the temperature indicated by the thermometer tj. It was then 

 passed through the tube Gi (which had a length of some fifty centimeters 

 inside the oil) into the dish Di. Sometimes the measurements were made 

 with the water in Di at rest, sometimes with the water flowing very 

 slowly from the tuhe, this giving a fresh surface as free as possible from 

 absorbed gases or contamination of any kind. 



The author feels sure of all his data except his temperature measure- 

 ments. The thermometers used were bought for high grade instruments. 

 It was the intention to calibrate them at the conclusion of the experiment. 

 By accident they were placed with some others of the same kind and so 

 could not be identified. 



The results obtained in this investigation are given in the following 

 table and are plotted in Fig. 3. 



Temperature Tension in 



of the Water — Dynes per cm. 



1.0° 



1.4 74.95 



6.6 .74.176 ' 



10.7 73.667 



16.5 73.087 



21.8 72.20 



29.2 70.795 



37.3 69.32 



50.4 67.36 



51.0 67.27 



61.6 65 . 50 



67.5 64.45 



72.6 63.71 



