256 Mr. J. A. Broun on the Annual [Mar. 26, 



The second instrument, made according to my own designs by Mr. P. 

 Adie, of London, had a magnet weighing only about one sixth of the other ; 

 it was suspended under a glass bell from which the air was exhausted, 

 and which was covered with two hoods one with gilt surfaces, the 

 other with cotton wadding. This instrument was placed in a closed room 

 without windows or external openings, and with a terraced ceiling below 

 the observatory roof. Observed from without (within the large room of 

 the observatory), the diurnal variations of temperature in the instrument 

 were not more than three tenths (0*3) of a degree Fahrenheit, while the 

 annual variation was under 5 Fahr.* 



The compared mean positions of the two magnets for each day, derived 

 from hourly observations of both instruments during eleven years, and 

 from eight daily observations during the remaining five years, will be 

 found with all other details in the volume referred to in the note to the 

 preceding paragraph. It will be sufficient for the purposes now in view 

 to give here the chief conclusions from these means. 



The monthly mean declinations having been freed from the secular 

 movement, and the means for three groups of years having been taken, 

 these means are represented very nearly by the following equations of 

 sines (6 = 0, Jan. 15) : 



Years, 



fAdie. 2/=0'-033 sin (0 + 135) + 0''069 sin (20+299). 

 1 Grubb.2/=0'-030 sm(0 + 150) + 0'-078sin(20+300). 



iftftftt 1ft4. /Adie. y=0'-190s 



: 1 Grubb. 2/=0'-099 sin (0 + 211) + 0'-062 sin (20 +319). 



J Adie ' ^=OH71 sin (0 + 181)+0'-104 sin (20+342). 

 1 Grubb. 2/=0'-062 sin (0+228) + 0'-122 sin (20+322). 



In the years 1854 to 1859 the movements of Grubb's telescope were 

 very small, the daily mean declinations from both instruments differing 

 rarely more than O''l throughout the whole six years. It will be seen 

 that the equations for these years agree very nearly. In spite of the 

 greater movements of the telescope in following years (affecting chiefly 

 the coefficient of sin 0), the epochs of maxima and minima derived from 

 the two instruments differ but little, and all the principal deviations 

 from the mean law for any year are confirmed by both instruments. 



When the means for the whole sixteen years are taken, and the equi- 



* Experiments with suspension-threads carrying slightly magnetic weights of nearly 

 one pound, showed that the effect of a change of 1 Fahr. on the position of Grubb's 

 magnet amounted to about 0''003 (=0"'18) a result deduced from the changes of tem- 

 perature from hour to hour, as well as from those from day to day. I must refer to 

 the first volume of the ' Trevandrum Observations,' now in the press, for the details of 

 these experiments. 



