724 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1912. 



temperatures on different sides were 38°, 37°, and 37°, at distances ranging from 130 

 to 1,320 feet from it. On that day the water temperature, on a line from Chateau Bay 

 to Belle Isle, was 36^° off the mouth of the bay, 39° in the middle, and 41° off the 

 south end of Belle Isle. It was lowered less than 2°, therefore, in the proximity of 

 the iceberg. 



The next day, August 8, a small iceberg was aground in Chateau Bay. The water 

 temperature in the middle of the bay was 34° and at the mouth 34^°. The lowest 

 temperature close to the iceberg was 33^°, which shows a difference of not more than 

 1° due to the iceberg. 



In 1906 an iceberg about 140 feet long was aground in 38 fathoms, about li miles 

 from Station P, where it remained for several days. On June 19 it was examined in a 

 boat. The surface temperature in the strait at the time was 35§°, and close around 

 the berg it was found to be the same, except on the west side, where the water tailing 

 from it with the flood was 35°. There was thus only one-half degree difference of tem- 

 perature to be found near it. 



It is clear that up to the time of the experiments with the micro- 

 thermometer, in 1910, there was good evidence to show that the 

 ordinary thermometer is useless to detect the small temperature 

 effect of an iceberg. Hence captains are correct in their statement 

 that the ship's thermometer is useless as a means for locating an 

 iceberg. 



THE RECORDING MICROTHERMOMETER. 



The development of the recording microthermometer has been 

 the result of nearly 20 years' experience in the study of minute 

 temperature changes in the ice-bearing water of the St. Lawrence 

 River. By appljdng very sensitive electrical-resistance thermometers, 

 it has been possible to show that the temperature of the St. Lawrence 

 in winter never varies more than a minute fraction of a degree from 

 the freezing point. The small variations that have been observed 

 and measured are the result of heat exchanges when ice is present, 

 and they accompany the formation or the disintegration of the ice. 

 The dehcate poising of the forces of nature are here wonderfully 

 illustrated. A few thousandths of a degree on either side of the 

 freezing point of the river water produces immense physical effects. 

 Thus the character of a river may be changed in a single night, or 

 the wheels of the largest hydroelectric station completely stopped by 

 a drop of a few thousandths of a degree in the temperature of the 

 water. As a result of this knowledge it is now possible to apply arti- 

 ficial heat around the wheels and gates, and completely prevent any 

 trouble from the sticking of the ice needles drawn in by the water. 



Four years ago I undertook some experiments to study the ice- 

 breaking operations on the St. Lawrence, and to determine the effect 

 of open water conditions on the temperature of the river. During 

 this time I turned my attention to a practical form of electrical- 

 resistance thermometer, which could not only easily measure thou- 

 sandths of a degree but automatically record them on a chart when 

 working from the ice breakers going at full speed. Following out 



