under others; the same is true of blue colloidal gold, which can be produced in a variety of ways, for example, by, the reduction of gold chloride solution by hydrazine. In most cases, however, the colloidal particles can not be seen even under the best conditions; they are, therefore, smaller than one seventh of a micron (1/7000 mm.), which is about the limit of microscopic visibility. It will be of interest to determine whether they can not be detected in many other cases with the help of the new Zeiss microscope, which, by employing quartz lenses and ultraviolet light (having a much shorter wave-length than ordinary light) and obtaining the image photographically, extends the limit of visible diameters to about one half of its present value. With such an 'ultramicroscope' a German investigator, Raehlmann, has already observed the suspended particles in an albumen solution. By the optical method of Sidentopf and Zsigmondy, in which the colloidal mixture is intensely illuminated by a thin beam of light, and the diffused light reflected from the suspended particles at right angles to the beam is viewed with a powerful microscope, the presence of still smaller particles having a diameter of 1/100 micron has been detected in red-gold suspensions and in other colloidal mixtures.
A strong indication of the heterogeneity of colloidal suspensions is furnished also by the familiar optical phenomenon, which is often called the Tyndall effect, and is observed when a beam of light is passed through any medium containing particles in suspension. The beam becomes visible, as does a sunbeam in dusty air, owing to a diffuse reflection of light from the particles. This can readily be shown to occur with colloidal suspensions of gold and of arsenious sulphide. Moreover, in every case where reflection takes place from non-metallic surfaces the reflected light is polarized, and this is found, in fact, to be true of the rays diffusely reflected from a colloidal suspension by examining them with a rotated Nicol prism. It has been shown, to be sure, that not only colloidal solutions (colloidal mixtures of the first class), but also ordinary solutions of some substances with complex molecules like sugar, exhibit this phenomenon, so that it is not a decisive criterion of a suspension. They do so, however, in an incomparably less degree than do typical colloidal suspensions, so that it at least furnishes evidence that the particles in the latter mixtures are of much larger size than are those in the former.
Whether the well-defined colloidal suspensions possess in appreciable degree what may well be regarded as the best single criterion of a true solution—a measurable osmotic pressure—does not, in spite of its importance, seem to have been the subject of investigation by the direct osmotic method. Nor is there conclusive evidence that they show the closely related phenomenon of diffusion. If the existence of these properties to an extent corresponding at all to the size and number of the particles should be demonstrated, it would, of course, prove
