J. 0. Graham — Experiments with an Artificial Geyser. 59 



Third, as to the specific heat. If the specific heat of the 

 soap solution were less than that of the water, less heat would 

 be required to bring a given amount to the boiling point, and 

 with a constant supply of heat, less time would be required. 

 This observation it was impossible for me to make with any 

 great accuracy, but from the results of my experiments, I 

 should say that there was no appreciable difference between 

 the two specific heats. 



Fourth, as to the retention of heat. In the tables concern- 

 ing the periods of eruption, it was noticed that the tempera- 

 ture was lowered more after an eruption in the case of the 

 soap and water than in the case of the pure water. As the 

 boiling point was the same in .both cases (compare Tables III 

 and IV) it was considered that this was due to the columns 

 above being less heated in one case than in the other. This at 

 once suggests a possible explanation. The heat is retained in 

 the lower part of the tube where it can be utilized in causing 

 the fluid to reach the boiling point, and so not wasted in rais- 

 ing the temperature of the whole column. How, then, is this 

 heat retained, or conversely, how is the heat lost in the case 

 of the water geyser ? 



Convection is, of course, the principal method by which 

 heat is conveyed from the lower part of the tube to the upper 

 portion. If the. viscosity of the fluid retards convection, this 

 then would cause the heat to be retained below. To test this 

 matter, I constructed a piece of apparatus bv which a column 

 of water was heated at the base only, and the temperature 

 attained by the water in the upper part of the tube could be 

 read at given intervals of time. By changing the fluid to 

 soap and water, the influence of the viscosity upon convection, 

 at least as far as it affected my problem, could be determined. 

 After a number of experiments with the fluids at high temper- 

 atures, I was somewhat surprised to find that the thermometer 

 in the soap solution showed in every case, a slightly greater 

 degree of heat than in the case of the water. That is, vis- 

 cosity did not seem to retard the escape of heat by convection. 

 Hence convection cannot be the factor sought. 



In all eruptions of the geyser the final out-rush of the water 

 is preceded by the rise of bubbles of steam (and of air, possi- 

 bly, to some extent), through the column of water. If these 

 bubbles are retarded in their ascent by the viscosity of the 

 fluid, they will give out more heat in the lower part of the 

 tube and so carry less to the upper. Accordingly, I devised 

 an apparatus to measure by the chronograph, the time required 

 for the bubbles to pass a given distance through the different 

 liquids against the force of gravity. The average of about 

 twenty five readings in each fluid showed that the time re- 



