[m'leod à BARNES] LOCAL TEMPERATURE FORECASTING 1S9 



was also passed through a length of lead tubing, was used to con- 

 nect the cage with the cable head box, and after repeated tests 

 of the insulation and zero of the instrument, the mountain thermome- 

 ter was put in its old place. Records have since been made in a per- 

 fectly satisfactory manner up to the time of writing, except for about 

 a week, when a thunderstorm during the middle of May produced 

 some slight damage. It is to be hoped that the records will now go 

 on without interruption. The first few days of every month are 

 devoted to bringing the mountain thermometer down to the Obser- 

 vatory and testing the zero of the instrument, which means tracing 

 on the record sheets the line of equal temperature. This is done, as 

 before described, by looping the mountain wires and connecting the 

 upper level thermometer into the circuit in the observatory cage, 

 side by side with the low level instrument. 



Our differential records during a part of January, and all of 

 February and March, are incorrect during a part of the day, owing to 

 the thermometer being enclosed in the cable head box. This box was 

 almost entirely closed to keep out rain and snow. As a result, the 

 sun, shining on the box and heating the interior, caused a large positive 

 departure of the record from the line of equal temperatures. No 

 abnormal effect was noticed at night or during dull cloudy days. 

 Nocturnal radiation may have produced some effect in cooling the 

 box, but this is small as compared with the effect of the sun's radia- 

 tion, and no abnormal negative departure of the trace was observed. 

 The monthly averages for February and March are therefore computed, 

 leaving out of account the effect of the sun during such time as that 

 was observed. So few traces were obtained in January that we have 

 not been able to give a value of the monthly average difference for 

 that month. 



The values from June to September, 1907, were taken solely from 

 the thermograph charts. The values for October, November and 

 December were from both thermograph and mountain instruments; 

 the thermograph was discontinued after December. It is interesting 

 to compare the differences obtained by the two methods during the 

 three end months of 1907. Thus the differences are as follows: — 



TABLE I. 



Month Difference by Same by 



1907 Thermograph Differential Recorder 



October — 3.°0 Fah. — 2.°9 Fah. 



November — 1.°0 Fah. — 2.°9 Fah. 



December — 3.°4 Fah. — 3.°6 Fah. 



