Notices respecting New Books. 153 



reductions and adjustments a table is ultimately obtained, showing 

 the results of the photometric determination of the comparative in- 

 tensities of the light of sixty-nine stars. The following are a few 

 of the results : — 



Sirius 4-052 



Canopus 1*994 



a Centauri 1 -000 



Arcturus 0*726 



Rigel 0-654 



Procyon 0-520 



a Orionis .... - 484 



Lyra 0*446 



Antares 0*404 



jS Centauri 0*399 



x Crucis 0*377 



a Aquihe 0'350 



Spica 0*309 



ij Argus 0*262 



Fomalhaut 0*262 



y Orionis 0*207 



Sir John then enters into a comparison of the values of the coeffi- 

 cient jjl, computed in the manner above explained, with those of M, 

 which represents the absolute quantity of moonlight, and on which the 

 former necessarily depend as their principal argument. The conclusion 

 at which he arrives is, that the value of jx is as the cube of the quantity 

 of moonlight. He also deduces this inference, namely, that " within 

 the limits of the experiments the effective impression of a star on the 

 retina is inversely as the square of the illumination of the ground of 

 the sky on which it is seen projected." 



The present chapter concludes with a comparison of the photo- 

 metric results with the conventional scale of naked eye magnitudes 

 adopted in the sequences. For this purpose the two series of values 

 (representing the conventional magnitudes on the one hand and the 

 light on the other) are projected as the abscissa and ordinates of an 

 interpolating curve ; then, calling x the conventional magnitude of 

 any star in both series, and y its light, the values of y corresponding 

 to every decimal of a magnitude from x—\ to ,r=3-6 are read off 

 from the curve. This curve is found to approach very nearly to a 

 cubic hyperbola, the equation of which is accordingly assumed ; and 

 the two constants which enter into the equation are determined from 

 particular values of x. Having thus an equation between the mag- 

 nitude and the light, the magnitudes of the several stars whose light 

 has been determined by the photometric observations are computed, 

 and the resulting values compared with those assigned by the se- 

 quence observations. A table is given, showing the magnitudes, as 

 determined by both methods, of sixty-eight stars ; but of these, the 

 five largest, — Sirius, Canopus, a Centauri, Arcturus, Lyra — should 

 be excluded from the comparison, the sequence magnitudes assigned 

 to them being purely arbitrary. Out of the remaining sixty-three 

 there are eighteen in which the disagreement of the assigned values 

 exceeds one-fifth of a magnitude (a quantity respecting which, it is 

 observed, the judgement of a practised eye cannot be deceived) ; but 

 on the other hand there are forty in respect of which the results agree 

 within an eighth of a magnitude. In respect of the cases of consi- 

 derable discordance, it is to be remarked, however, that in a majority 

 of them the photometric determination rests on a single night's ob- 

 servation, and not unfrequently on a single equalisation. Subjoined 



