harper: thermometry lag 205 



The process of obtaining the data is tedious but the advantage 

 is offered of a minimum of interference with the current to be 

 measured. As an example of its application may be mentioned 

 a determination of the flue draft of a continuous flow combustion 

 calorimeter of the Junkers type, where it is evident that none of 

 the more usual forms of anemometer could be used. 



Lag of a Beckmann thermometer. — While the ordinary "chemi- 

 cal" thermometer was found to operate in close accord with the 

 behavior predicted by equation (1), this was not the case for a 

 thermometer with part of its bulb in good thermal contact with 

 the immersion bath and part in very poor thermal contact owing 

 to enclosure in a surrounding air jacket. The usual form of Beck- 

 mann instrument has a large capillary between the bulb and the 

 fine capillary extending along the scale, and this large capillary, 

 containing from 0.2 per cent to 0.5 per cent of the total mercury, 

 is to be regarded as a secondary bulb of appreciable influence. 



Assuming that equation (1) represents the behavior of the main 

 bulb, with a lag X B , and that also such an equation represents 

 the behavior of the large capillary, with a lag X c (much greater 

 than X b ) we can develop a theory analogous to the simpler one, 

 but containing two Xs. The equations are of course more com- 

 plicated. For instance 



^*-° = (1- k) e'K 1 +k<r\' (6) 



O — u 



replaces (4). k is the fraction of the total mercury contained in 

 the large capillary. Similarly (5) is replaced by 



u - 6 = (1 - k) r\ B + kr\ c (7) 



Methods of determining X B and X c were devised, and the actual 

 behavior of a Beckmann thermometer, when plunged into a bath 

 maintained at constant temperature, was found to be very closely 

 represented by equation (6) ; in marked contrast to the failure 

 of equation (4) in this respect. The assumptions made seem 

 therefore to be justified and the theory based on them is offered 

 as a better representation of the behavior of such a thermometer 

 than is the simpler theory employing but one lag. 



