xiv 



the positions lie nearly in one straight line, and from the green to the 

 blue in another, the chief divergence being at the red and blue ends. 

 The conclusion is that there are three primary colours in the spectrum, 

 red, green, and blue, by mixture of which colours chromatically 

 identical with the other colours of the spectrum can be obtained. The 

 position of the green is one-fourth from the line E towards the line F, 

 but the red and blue cannot be satisfactorily placed. Young's theory 

 was examined by Helmholtz as well as by Maxwell, with the like result, 

 which was to confirm its truth. 



If we pass to other subjects we must regard the essay on Saturn's 

 Rings as a most valuable investigation on account of the important 

 results it contains. Some of these we will now mention. The author 

 first establishes that the stability of a uniform solid ring revolving 

 round the planet is impossible. If we suppose the ring loaded at a 

 point of its circumference a possible case of stability is obtained, but 

 the load must be between - 8158 and *8279 of the whole mass of the 

 ring. So irregular a distribution of the mass is very improbable, 

 and would be found by observation. When, in addition, we con- 

 sider the immense size of the rings — an iron ring of such a size would 

 be exceedingly plastic under the forces it would experience, — when we 

 consider also their comparative thinness, it is impossible to regard 

 them as rigid. The stability of a ring of equal satellites is then 

 examined. Disturbances perpendicular to the plane of such a ring 

 cannot produce instability. In the plane of the ring the greatest 

 danger to stability will exist when the number of undulations of the 

 circle of satellites supposed oscillating about their mean positions is 

 the greatest possible, that is, when equal to half the number of satel- 

 lites. In that case, for stability, the mass of the planet must be 

 greater than *4352 / a 2 times the mass of the ring, where fi is the 

 number of satellites. If this condition hold, four distinct oscillations 

 of different periods and amplitudes will run round the circle of satel- 

 lites. Results similar in their general outlines, though necessarily 

 characterised by greater complexity, hold in the case of a ring of 

 unequal satellites. The next hypothesis examined is that of an 

 annular cloud of meteoric stones revolving uniformly about the 

 planet. It is shown that the average density of such a cloud must 

 be less than 300 times that of the planet, otherwise destructive 

 oscillations will be set* up in the ring. 



So low a density has been shown by Laplace to be impossible in a 

 ring revolving as a whole, so that the outer and inner portions cannot 

 have the same angular velocity. 



The final conclusion is that the rings of Saturn are composed 

 of an indefinite number of free particles revolving round the planet 

 with velocities depending on their distances from his centre. The 

 particles may be arranged in a series of concentric rings, or they may 



