抄録
Because of the recent progress of instruments, especially the thermistor, we can now easily make observations at more than 80 stations in an hour. However, serious errors are sometimes caused by the time lag of the thermistor in rapidly moving automobiles. The present paper deals with the errors in several cases, theoretically and experimentally, and following conclusions are summarized:
1) For using thermistor thermometer, with a time lag coefficient of about 3 seconds, the first reading in the automobile should be taken after having run at least 100 meters.
2) The influence of the thermal frontal zone between the urban and rural area, is experienced on both sides of the zone for 20 meters width.
3) When the automobile moves on a route where a temperature gradient exists, observed values are frequently overestimated or underestimated by 0.1_??_0.2°C.
4) Therefore when the two automobiles cross the same location, the difference between the values is 0.1°_??_0.2°C.
5) An extraordinary high temperature is caused by exaust gas from other automobiles, so no observation should be taken permitted within 50 meters at least, even if the influence has disappeared. Also same limitations must be considered on the observations made at the stopping at the crossing point.
6) The observed values at each station seem to be influenced by the area 50 to 100 meters before it.
7) In the moving observation from automobiles, the observed values fluctuate less than at the fixed stations, hence it is enough to consider the fluctuations only for the longer periods.
