Interference - Apparatus for Extenso meters. 129 



of the four thermometers indicated correctly the temperatures 

 of their bulbs. But this assumption is seldom valid in the 

 case of thermometers such as those used by students and 

 costing about 3s. 6d. each. The error due to this imper- 

 fection has the greatest influence in the case of the thermo- 

 meters C and D, since # 3 — # 4 is much less than ^ — 6 2 . 

 An approximate method of obtaining the corrections to be 

 applied to 6 x — 6. 2 and to #3 — #4 is to compare A and B in a 

 bath whose temperature is roughly i(0: + ^) and to compare 

 C and D at the temperature j>(0a-\-6±). When this was done 

 in the above experiment the results became 



(a) l -0 2 =36°-2C. J 3 -0 4 =lO°-9C., K = -91I. 

 (J,) 0,-0 2 =38 o -9C., 3 -0 4 =4°-8C., K = -$88. 



As the thermometers were only read by estimation to 

 1 1 (l degree, the agreement between these two values of K 

 is as close as could be expected. 



When sufficient time is available, the student should com- 

 pare the readings of each of the four thermometers with 

 those of a standard thermometer, whose corrections have 

 been determined at the National Physical Laboratory, by 

 placing them together in a bath of water which is gradually 

 heated. The whole of the standard thermometer must be im- 

 mersed in the bath, but as to each of the others, so much 

 of the stem should be in the siir as was the case when it was 

 fixed in the apparatus. In this way the correction due to 

 the u emergent column"" is practically avoided. The first- 

 hand knowledge of the limitations of thermometry (with 

 cheap thermometers) so obtained is well worth the time 

 -pent upon the comparison. 



Cavendish Laboratory, 

 Cambridge. 



X. An Interference Apparatus for the Calibration of Extenso- 

 meters. By John Morrow, 3I.Sc. (Vict.), M.Eng. 

 (Liverpool), and Ernest L. Watkix, M.A. (Cantab.), 

 Lecturers in University College, Bristol *. 



Introduction. 



BEING engaged in a research which requires the use 

 of extensometers of considerable sensitiveness, the 

 authors have found some difficulty in determining the con- 

 stants of the instruments with sufficient accuracy ; thev also 



* Communicated by the Physical Society : read October 28, 1904. 

 Phil. Ma : ,. S. H. Vol. 9. No. 49. Jan. 1905. K 



