380 ASTRONOMICAL OBSERVATIONS MADE AT [47 



positive or negative. The apparent Zenith distance thus obtained is used 

 with the data in the three next columns for the calculation of refraction. 



The thirteenth column contains the height of the barometer, as shewn 

 by a cistern-barometer constructed by Dollond and attached to the Circle 

 pier. The lower surface of the mercury is raised by a screw pressing the 

 bag till the light seen below a brass edge is excluded ; and a brass 

 slider is brought to the upper surface to shut out the light in the 

 same way. 



Before calculating the refraction, a correction of +0'01in. was applied 

 to these Barometer-readings [see Introduction to Vol. xx., p. cxvi.] for 

 Index-error ; but a comparison with a very fine Standard Barometer by 

 Adie, which was mounted in the Transit Room in July, 1872, seems to 

 shew that this correction is too small. A large number of comparisons 

 made between August, 1872, and the end of the year, shew that the 

 reading of Adie's Barometer exceeds that of Dollond's by 0'055 in., and 

 the correction of Adie's Barometer, by comparisons with the Standard 

 Barometer at Kew, is only O'OOl in. Probably the error of the old 

 Barometer had been gradually increasing. 



The fourteenth column contains the reading of the thermometer whose 

 bulb is plunged in the cistern of the barometer. 



The fifteenth column contains the reading of an external thermometer, 

 which is fixed to a stage near the north shutter-opening at a distance 

 of four feet from the wall of the building and nine feet from the ground. 

 It is protected from radiation and from the weather, and contiguous 

 parts of the building prevent the direct rays of the Sun from falling 

 upon it. 



The refraction is calculated by Bessel's Tables, using the convenient 

 form in which they are given in the Appendix to the Greenwich Obser- 

 vations for 1836. In this mode of calculation the reading of the attached 

 is supposed to be the same as that of the external thermometer. The 

 former reading, though not made use of, is inserted in the printed 

 columns, to allow of correcting for the error of this supposition, if it is 

 thought necessary. 



By adding the refraction to the apparent Zenith distance North or 

 South, the true Zenith distance is found, and by adding algebraically 

 the true Zenith distance, considered negative when north of the Zenith, 

 to the assumed co-latitude of the Observatory, viz. 37 47' 8"'00, the 



