1. Get a correct idea of the different characters of the primary clouds. The true fibrous _cirrus_--the different forms of _cirro-stratus_--the smooth, uniform _stratus_--the _cirro-cumulus_, which is nothing but a cirro-stratus, separated into _distinct masses_ by the repulsion of static electricity--and the _cumulus_, too distinct ever to be mistaken. There is no difficulty, except with the varied forms of cirro-stratus. It is useless to attempt to give, or the observer to rely on, names for these numerous forms, without as numerous illustrations. Those in use are rarely applied correctly. I have never met with ten persons who applied even the term "mackerel sky" to the same precise form of cirro-stratus. In relation to all of them it is to be observed that polar belts of condensation, and local appearances of considerable extent, are often too feeble in action to precipitate, even when the mackerel form is present; and all may be the lateral attendants of passing storms. Therefore, 2. Satisfy yourself whether the cirrus or cirro-stratus increases in density and tends to the formation, or induction, of stratus; and whether it is isolated, or an extension of the condensation of a storm, and if the latter, _where that storm is_. The time will come when an intelligent use of the telegraph will do this for you. 3. Look also to the character of the wind, if there be any. On this subject I have perhaps said all that is necessary in the preceding pages. Next to condensation, the direction and character of the wind is the most valuable prognostic. Indeed it often tells us that a storm is approaching, and the quarter from which it will come, and its character, before the condensation is visible. 4. See if there is any _secondary_ condensation or scud. These are sometimes seen running toward a storm, when there are not distinct clouds visible in the western horizon, at nightfall, or in the evening, as in the instance stated in the introduction, and sometimes from the north-east, as in cases heretofore so often stated. But the easterly scud do not often form in winter, until after the cirrus has passed into the form of cirro-stratus, or has induced the latter forms in the inferior portion of the trade, or the surface atmosphere.