There is also another instrument for detecting polarization which is perhaps more commonly used than the biquartz: it is generally called Savart’s analyser, and is extremely sensitive in its action. On looking through it at any object emitting ordinary light, the white circle of light limited by the aperture of the instrument only is seen; but if any polarized light should happen to be present, a number of parallel bands, each shaded from red to violet, make their appearance; on rotating the instrument a point is found when a very slight motion causes the bands to vanish and others to appear in the intermediate spaces, and knowing the position required for the change of bands with light polarized in a known plane, say the vertical plane, it is easy to find how far the plane of polarization of any ray is from the vertical, by the number of degrees through which the instrument must be turned to change the bands. The construction of the instrument, and especially its action, is not easy to understand without a considerable knowledge of optics, but it may be stated that a plate of quartz is cut, in a direction inclined at 45° to its axis, into two parts of the same thickness; one part is then turned through a right angle and placed with the same surfaces in contact as before; these are fixed in the instrument so that the light shall traverse them perpendicularly to the plane of section; the light then passes through a Nicol’s prism as an analyser to the eye. The lines observed, “black centred” in one position, and “white centred” in the position at right angles to this, are always in the direction before referred to. The delicacy of the test supplied by this arrangement increases as this direction is more nearly parallel or perpendicular to the plane of polarization of the ray under examination. CHAPTER XXXI. CELESTIAL PHOTOGRAPHY.—THE WAYS AND MEANS. We come now last of all to that branch of the work of the physical astronomer which bids fair in the future to replace all existing methods of observation. In the introductory chapter we referred to the introduction of photographic records of astronomical phenomena as marking an epoch in the development of the science. In the last ones we have to dwell briefly on the _modus operandi_ of the various methods by which the eye is thus being gradually replaced. The point of celestial photography is that it not only enables us to determine form and place, absolutely irrespective of personal equation so far as the eye is concerned, but that, properly done, it gives us a faithful and lasting record of the operation, so that it is not forgotten; Mr. De La Rue has called the photographic plate the _retina which does not forget_, and an excellent name it is.