384 



HENRY A. BOWLAND 



In using these tables a correction is of course to be made should the 

 zero point change. 



TABLE XXI. CORRECTION OF KEW STANDARD TO THE ABSOLUTE SCALE. 



Appendix to Thermometry 



The last of January, 1879, Mr. S. W. Holman, of the Massachusetts 

 Institute of Technology, came to Baltimore to compare some thermom- 

 eters with the air thermometer; and by his kindness I will give here 

 the results of the comparison which we then made together. 



As in this comparison some thermometers made by Fastre in 1851 

 were used, the results are of the greatest interest. 



The tables are calculated with the newest value for the coefficient of 

 expansion of glass. The calibration of all the thermometers, except 

 the two by Casella, has been examined, and found good. The Casella 

 thermometers had no reservoir at the top, and could not thus be readily 

 calibrated after being made. The G-eissler also had none, but I suc- 

 ceeded in separating a column. 



The absence of a reservoir at the top should immediately condemn 

 a standard, for there is no certainty in the work done with it. 



From these tables we would draw the inference that No. 6163 repre- 

 sents the air thermometer with considerable accuracy. At the same 

 time, both tables would give a smaller value of ra than I have used, 

 and not very far from the value found before by direct comparison, 

 namely, -00015. 



The difference from using m= -00018 would be a little over 0-01 C. 

 at the 20 point. 



All the other thermometers stand above the air thermometer, between 

 and 100, by amounts ranging between about 0-05 and 0-35C., 



