Notices respecting New Booh. 151 



carries the prism and lens, and the requisite motion is given by 

 simple contrivances, into the details of which it is unnecessary 

 to enter. In using the instrument, a star must be chosen not more 

 than 102° 43', nor less than 60° or 70° from the moon, which ought 

 not to be too near the full. The staff being first directed to the star 

 is turned on its axis till the prism is brought into the proper position ; 

 and when the proper angle of reflexion is attained, the artificial star 

 (or moon) is seen side by side with the real star under comparison. 

 The slider is then approached to or removed from the eye, along the 

 staff, until a perfect equality between the real and artificial star is 

 obtained, when the distance of the slider from the eye is measured 

 by means of a measuring -tape connected with the apparatus. A 

 sufficient number of equalisations ought to be made to destroy the 

 numerous sources of illusion to which the judgement of equality is 

 subject, and give confidence in a mean of the readings. This done, 

 the other star is treated in the same manner. 



Sir John states that the actual comparisons made with the instru- 

 ment were neither very numerous nor performed under the most fa- 

 vourable circumstances. In several of them the moon was nearly full ; 

 and though he had expected this to influence the results, he was by 

 no means prepared for the enormous extent of that influence. In the 

 original conception of the process, he had hoped that the image of 

 the moon, when reduced to that of a mean full moon, by allowing 

 for the phase and apparent diameter, would afford a definite standard 

 of light ; but the effect of the illumination of the sky destroys all 

 possibility of obtaining absolute results in this manner, and renders 

 it necessary to regard them as comparable only with those taken on 

 the same night and under the same circumstances. All these diffi- 

 culties and objections, he adds, might be got over by using Jupiter 

 instead of the moon as a standard luminary ; but this idea did not 

 occur to him when it might have been serviceable. He refers also 

 to an instrument constructed on the same principle, — that is to say, 

 in which a prism is made use of to deflect the light of a standard 

 luminary by total reflexion at its base, — invented by M. Steinheil, 

 and communicated to the Gottingen Academy in 1835 j but he had 

 no knowledge of the existence of M. Steinheil's memoir till his return 

 to Europe in 1838. 



The observations with the astrometer were made in the year 1836 

 on nineteen different days, between March 28 and December 26. In 

 each case the numerical result is obtained from the following formula : 

 " Let A represent the moon's augmented semidiameter, in seconds, at 

 the place and moment of observation as given by the Nautical 

 Almanac and a table of augmentations ; e her elongation from the 

 sun, R the earth's radius vector, the mean value of R being =1 ; 

 933"'5 the moon's apparent diameter. Take 



M = 1000 / —, — sin -V, 



1#X933"'5 2 J 



then will M express the absolute quantity of light in the moon's 

 image (supposing all her illuminated surface equally bright) and 



