2S6 BELL SYSTEM TECHNICAL JOURNAL 



bridge," thus recording the values of humidity directly. Inasmuch as 

 successive operations of the recorder consist in the restoration of the 

 balance of each bridge, if different from the last position of balance, it 

 is evident that the pens will trace continuousK- the \arialions of 

 temperature and relali\e humiility. 



A recorder of this type with its associated wind tunnel mechanism 

 has been used for some time to record the conditions in a laboratory 

 "humidity room." The temperature record gi\'en by this recorder is 

 accurate to ±i.t°F. as in the case of the difference recorder. Tlio 

 accuracy of the humidity record differs for various points on the scale, 

 depending upon the values chosen for certain resistances in the recorder. 

 When the recorder is adjusted for very close accuracy (±3-2% relati\'e 

 humidity) for relali\e humidities above 90%, the accuracy for lower 

 values of humidity decreases until at 50% the maximum \ariation 

 from the true value may be as much as 2y>% relatixe humidity. 

 If desired, the adjustment ma\' be made to transfer the jioinl of 

 greatest accuracy to any selected lower \aliie of humidity. Experi- 

 ence with this model has suggested changes which should considerably 

 improve this accuracy over the whole range of humidities. Fig. 12 

 shows a typical 12 hour record of conditions in the "humitlil>- room" 

 while under automatic control of an air conditioning e(iui|5meni. 



This recorder also was used during the summer months of lil23 

 to record outdoor conditions with the wind tunnel equipment installed 

 in the Weather Bureau instrument shelter mentioned earlier. During 

 this period of 4 months' operation, it required no attention sa\'e an 

 occasional oiling of the mechanism and maintenance of thi' wi-t l)u]l) 

 equipment, and |)ractically continuous records were sciuied. I'lie 

 records are of particular interest for observation of the xarialions of 

 temperature and humidity which take place during cJianges in 

 weather conditions such as rain storms. Figs. 1,S and 14 are 

 reproeluctions of typical consecutive 12 hour recf)rds obtained for 

 outdoor conditions. 



I'roni consitleralioii of ihc iiuniiditx- recording apparatus wiiicli lias 

 been developed and i he rouits wliiili ha\e been obtained with ii . ii nia\ 

 be stated that both the difference recorder and the direil re.iding 

 recorder are satisfactory instruments with which accurate data may 

 be obtained. However, the\' are instruments which, in common with 

 other types of apparatus that lui\e been dewioped to measure 

 humidit\-, require careful alicnlion of the wind timnel etiuipiiKMit 

 in order to secure reliable roiills; also the recorder mechanism itself 

 reciuircs the attention of an opiTator skilled in its maintenance. 



