48 Mr. J. B. N. Hennessey. [May 25, 



glass off; for those of No. 2 the unit is o, 01 F., assuming a 

 graduation of No. 2 to be 0°'2 F. 



8. It needs but little familiarity with the two kinds of instruments 

 -to decide in favour of that by Professor Balfour Stewart over that by 

 the Rev. G. C. Hodgkinson, a preference which may be greatly in- 

 creased by doubling the sensitiveness of the former. I have accord- 

 ingly converted all my results beginning with 1869, and expressed 

 them in terms of Stewart's No. 2, or 0°"01 F., which unit is to be 

 understood in all that now follows. 



9. Again, as the average zenith of distance of the sun during my 

 autumn series does not differ greatly one year with another, and as 

 suitable corrections to an adopted zenith distance may be inferred 

 from the long range series {i.e., three or four hours on each side of 

 the meridian) taken at the same season of the year, I have reduced 

 all my autumn results to the zenith distance of 45°. 



10. The long range series used are given in T. 4, where it will be 

 -seen that the series with A in 1881 is inconsistent, so that I have 

 rejected it. It appears a waste of time to attempt formulating at 

 present the correction from one Z.D. to another ; I made several 

 endeavours of the kind with the long series of T. 4, but feeling that 

 the expressions were more pretentious than real, I resorted to the 

 projection shown in the correction Curve C 1 : here it will be seen 

 that the two curves, i.e., of 1879 and 1881, are fairly similar, and as 

 the series of 1881 was taken at half-hour intervals instead of the one- 

 hour intervals of 1879, the correction curve I have adopted is made 

 to agree more nearly with the later series. All the reductions to 45° 

 Z.D. have been made with the help of C 1. 



11. From T. 4, long series, rejecting the results of A, 1881, we 

 have — 



App. time. Defect from noon. 



h. 1879. 1881. 



8 115 



9 45 39 



9.5 26 



10 19 15 



10. 5 6 



11 6 4 



1 15 6 



1.5 11 . 



2 30 , 18 



2. 5 25 



3 62 55 



4 127 



where the 1879 series is hourly, and from 8 a.m. to 4 p.m., while the 



