.s7A- \VII.UA.W .sv/-:.i//-:.v\, I.R.S. 271 



During the above experiments the sea was perfectly smooth, 

 with no wind. The ship's engines were not used at all, the vessel 

 lying almost motionless in the water. The temperature of the 

 comparison coil was reduced by water from a carafe, the wat-T 

 contained therein being frozen by a Carre" ice machine. Two 

 carafes were prepared at a time, and there was plenty of time to 

 keep one constantly at hand. 



In order to allow the Miller-Casella thermometers to record the 

 high temperature of 50 fathoms in the last series, they were 

 lowered very rapidly to that depth, and after eight minutes reeled 

 back at the rate of 200 fathoms per minute, so that the minimum 

 side had not time to assume a lower temperature. 



The cable was led from a large reel through an 18-inch leading 

 block, and was. lowered and reeled in very slowly, and without 

 jerks. 



It may be noted in the above Tables that the two instruments 

 gave precisely the same readings at positions of maximum or 

 minimum temperature, but that in intermediate positions the 

 electrical thermometer, in almost every instance, gave a higher 

 reading. This discrepancy may be accounted for, I think, by the 

 circumstance that the electrical thermometer gives the tempera- 

 ture of the water actually surrounding the coil at the moment of 

 observation, whereas the reading of the Miller-Casella instrument 

 must be affected by the maximum or minimum temperatures 

 encountered in its ascent or descent, which may not coincide with 

 that at the points of stoppage. A strong argument in favour of 

 the electrical instrument for geodetic and meteorological purposes 

 has thus been furnished. 



