The prismatic play of color which this gem alone possesses to any considerable degree constitutes its chief charm, and its cause has been a matter of earnest study among opticians. A plausible theory has lately been advanced by an English philosopher that the colored rays are produced by the relation of the high refractive to its very low dispersive power. For instance, this refractive power in the diamond, or, in other words, its property of bending a ray of light falling obliquely upon its surface, is 2.439, while that of water is only 1.336, and that of glass 1.500. But its power of dispersing a ray of white light, or, in other words, of separating it into its compound colors in reference to its refractive power, is only 0.038, while that of glass is 0.052. Hence it is surmised that this inferiority of dispersive power is required for the production of the splendid colored reflections which constitute the glory of the gem. It is also maintained that this high refractive power separates the red and the blue rays more than a high dispersive power would in other transparent bodies, and to such degree as to allow each color of the spectrum its full force. As example, the zircon, with its inferior reflections, is offered, its refraction being 1.99 on the established scale, while its dispersive power is as high as 0.044. The relations of the spinel are also as 1.81 to 0.040, and neither does the gem display the rainbow hues. This theory is certainly ingenious, and if correct the test may be applied to other transparent minerals possessing similar relations. We may, therefore, expect the white garnet to exhibit the property of prismatic display, as it has a refractive power of 1.81 and a dispersive power of 0.033. But, unfortunately, perfectly pure and transparent white garnets are unknown, and we must therefore turn to other minerals for comparison. To the white tourmaline, then, we will apply the test, since this mineral has a refractive power of 1.66, with a dispersive power of only 0.028. Here, then, we have nearly the same relation as observed in the diamond; and, if the theory be correct, we may reasonably expect the exhibition of the same phenomena. But, upon examination of several perfectly white and transparent tourmalines from Mt. Mica, cut into regular brilliants, we have failed to detect an increase of prismatic display, or even discover any evidence to lend support to the plausibility of the hypothesis. We, therefore, reluctantly turn to other arguments for a solution of this most interesting problem.