48 



KNOWLEDGE 



[March 2, 1891. 



nical letters and days of the month should be mounted 

 on a circular piece of cardboard, affixed by a paper- 

 fastener or button and thread through its centre to a 

 larger piece of cardboard, on which the outer circle con- 





JULY 







%8 



o. 



.4^ 







\rK^<'' 



I 







taiuiiig the days of the weeli and months should be 

 gummed concentrically with the inner disc. 



Mr. Prince's rules for finding the day of the week 

 A List of Sunday Letters Corresponding to any date in this 

 from A. D, iSooto iSqq. ccntury, are as follows : — 



" Kotate the Circular Index until you 

 place the Sunday Letter for the Ye.ui 

 under the Month you require, when 

 it will show the day of the week for 

 any day in that month, and will thus 

 serve as a Calendar for any number 

 of years. The Letters in the first 

 column of the adjoining Table are 

 the Sunday Letters for each year in 

 the same line." 



" Li the case of Leap Year (to Feb- 

 ruary 29th only), make use of the 

 first of the two Sunday Letters for 

 that year ; thus, for the year 1888, A 

 woidd be the letter to February 29th, 

 and G for the remainder of the year. " 

 "Each Letter in sixth column de- 

 notes that all years between it and 

 that in column one are Leap Years. 

 This table renders the Calendar use- 

 ful for ascertaining the day of the 

 week of any date during the present 

 century." 



The days of the week for next century may be at once 

 determined by finding the day of the week for the corre- 

 sponding day of this century, and gomg back two days ; 

 thus, the 1st of January 1801 was on a Thursday, and the 

 1st of .January 1901 will be on a Tuesday, The days of 

 the week for last century, from the 1-lth of September 

 1752, when the eleven days were dropped, till the end of 

 the century, may be determined by finding the day of the 

 week for the corresponding day of this century and going 

 forward two days ; thus, the 1st of January 1853 was 

 on a Saturday, and the 1st of January 1753 was on a 

 Monday. 



To determuie the days of the week for the twenty-first 



century we must go backwards three days in the week 

 from the corresponding day of this century, for the year 

 2000 will be a leap year, while the years 1900, 2100, &c. 

 are not leap years. For the twenty-second century the 

 rule will be — go backwards five days in the week, which is 

 the same thing as going forward two days. 



In the twenty-third century, people will keep the tetra- 

 centenaries, or fourth-century celebrations of events which 

 have happened in this century, on the same day of the 

 week as that on which they actually happened or are 

 suj)posed to have happened. 



In future all tetracentenaries will occur on the same 

 day of the week as the event they commemorate, for a 

 period of 400 years must always include one year which 

 will divide by 400 without leaving a remainder. 



NOTE ON THE ORBIT OF THE DOUBLE 

 STAR 5 2. 



By S. W. Buknham. 



IT would have been hardly possible to get even an 

 approximate orbit of this star with the measures 

 made during the fifty years following its discovery. 

 At the time of the first observations by Struve it 

 was comparatively easy to measure ; but the stars 

 slowly approached each other, and after 1858 the measures 

 are few and very imcertain. Many of the estimated, or 

 partly measured position-angles are obviously very largely 

 in error, judging from the earlier measures which may be 

 assumed to be reasonably exact, since the distance between 



the components was sufficient to make it easily measure- 

 able with a moderate aperture. No attempt, so far as I 

 am aware, has been made to compute an orbit from these 

 measures. The entire angular motion from 1830 to 1858 

 was only 12^. The principal change was in the distance, 

 which at the last-named date was only one-half that at 

 the time of the first measures by Struve. So far as these 

 observations are concerned, they coidd be as well repre- 



