A. Upper clouds. (_a_) Cirrus. (_b_) Cirro-stratus. B. Intermediate clouds. (_a_) Cirro-cumulus and alto-cumulus. (_b_) Alto-stratus. C. Lower clouds. (_a_) Strato-cumulus. (_b_) Nimbus. D. Clouds of diurnal ascending currents. (_a_) Cumulus and cumulo-nimbus. E. High fogs. (_b_) Stratus. In this tabulation the forms marked (_a_) are detached and occur in dry weather, while those marked (_b_) are widely extended. The original scheme also gives the mean heights of the various types, but these values have been omitted here because they are extremely variable, and impossible to ascertain with any approach to accuracy by mere eye estimates. They vary also with the season, and probably also with the locality. Moreover, the altitude is no guide to the name, except that on the whole the types occur in the order given, taking group A as the highest and group E as the lowest. In the chapter on cloud altitudes this subject will be further considered, and under the descriptions of cloud-forms their average height or actual measurements for the particular specimen figured will be given whenever possible. Before coming to the description of individual forms, it may not be out of place to give brief consideration to the best means of observing them in nature. For eye observation, of course, no directions are needed when we are dealing with the lower and denser varieties; but when we come to the highest groups it sometimes becomes necessary to protect the eye from the brilliant glare which may make it impossible to detect the real structure. Smoked glass, neutral-tinted spectacles, or yellow glass all have something to recommend them; but by far the most convenient means is to look, not at the clouds themselves, but at their images formed in a black mirror. A lantern cover glass, or a thin piece of plate-glass, blacked on the back with some black paint, serves admirably. But all black paints are not equally good. The best are oil paints which dry with a glossy surface, the so-called enamels. They have the advantage that the varnish with which they are mixed has an index of refraction not very different from that of the glass. The consequence is that so little light is reflected from the blackened back, compared with that which is reflected from the front surface of the glass, that the second image can only be detected with difficulty. If the mirror is a piece of black or deeply coloured glass all trace of the second image is lost.