Reading the Sky
Clouds had no names until a Quaker pharmacist gave them Latin ones, and the names turned out to be a forecast.

Say the sky this morning is clear except for a few white hooks drawn very high, curved at one end and combed all the same way. There is a decent chance it will be raining here tomorrow afternoon, and you could have said so before breakfast without looking at a phone. Those hooks are cirrus, they sit between twenty and forty thousand feet, they are made entirely of ice, and they are the leading edge of a mass of warm air sliding up over the colder air you are standing in. What follows them follows in a fixed order, and learning that order is most of what weathercraft amounts to.
For almost all of human history nobody could put that sentence together, because the sky had no working vocabulary. On a cold evening in December 1802, in the basement of the Plough Court pharmacy off Lombard Street in London, a thirty-year-old Quaker chemist named Luke Howard read a paper to his scientific debating club, the Askesian Society. He had kept weather journals since he was ten. The settled view was that clouds shifted too fast to be classified at all, that they were passing shapes and not kinds of thing. Howard gave the fundamental forms Latin names on the model already used for plants and animals: stratus for a layer, cumulus for a heap, cirrus for a lock of hair, nimbus for the cloud that rains. They could be combined, so that cirro-stratus named exactly what it looked like. The paper reached the Philosophical Magazine the following year and made him, to his considerable discomfort, a celebrity.
Before he could name anything he had to argue that there was something there to name, and he did it by conceding the strongest form of the objection first.
IIf Clouds were the mere result of the condensation of in the masses of atmosphere which they occupy … then indeed might the study of them be deemed .
IIThe principal are commonly as distinguishable from each other as a Tree from a Hill, or the latter from a Lake.
IIIStratus: a widely extended, continuous, horizontal sheet, increasing from below. Cumulus: convex or conical heaps, increasing upward from a horizontal base. Cirrus: parallel, , or diverging fibres, extensible in any or in all directions.
IVA correct comprehension of the subject … is only to be obtained by a habitual observation of Nature. [He is warning the young against leaning on the pictures.]
Luke Howard, “On the Modifications of Clouds,” read to the Askesian Society in December 1802 and published in 1803. Passages I, II and IV as quoted by the Public Domain Review from the 1865 third edition; III as quoted by the Cloud Appreciation Society. The capitals and the bracketed frame are as given by those sources.
The essay travelled. A German translation reached Goethe in 1815, and he answered with a poem in Howard’s honour whose tribute stanza closes, “As clouds ascend, are folded, scatter, fall, / Let the world think of thee who taught it all.” Constable and Turner met the system in Thomas Forster’s Researches About Atmospheric Phaenomena, and both began making sky studies around 1820. Constable inscribed his with the date, the hour and the wind direction, a bookkeeping habit no painter needs.
The World Meteorological Organization keeps the official register today, and it is still Howard’s Latin. The first International Cloud Atlas appeared in 1896, and the current list runs to ten genera sorted into three levels by the height of the base. The 2017 edition, the first published digitally, added one species, volutus, the long horizontal roll cloud, and five supplementary features including asperitas, the wave-like underside amateurs had been photographing for years. It added no genus at all. Words proposed to a dozen friends in a cellar have needed no replacement in two centuries.
Howard named the forms. Why they arrive in a particular order was answered in a small country in the last year of a war. Germany had cut Norway off from continental weather telegrams and the harvests of 1917 and 1918 were bad enough that famine looked possible, so the physicist Vilhelm Bjerknes used the emergency to win government backing for a dense national net of observing stations, built for fishermen and farmers, reporting often enough to show a weather system in motion. The Bergen service began issuing daily forecasts in July 1918 with his son Jacob as chief forecaster. Late in 1919, arguing over the lines of converging wind on a day’s map, they concluded that one boundary might run most of the way around the hemisphere, and reached for a word from a war twelve months over. They called it a front, and its two edges the warm front and the cold.
That word carries the mechanism. Warm air is lighter, so where it meets a cold mass it rides up over the top instead of shouldering through, on a slope shallow enough that the ramp of warm air reaches cloud height hundreds of miles ahead of the point where the front touches the ground. Cloud forms the whole way up that climb, thin and icy at the top, thick and wet where it comes down to meet you. The cloud overhead is therefore a measurement of how far away the front still is, and what you watch through a day is the ramp being drawn over your head from its far end to its near one.
In practice you are watching four things in a row. Cirrus first, the hooks and streaks, high and detached and whiter than anything else in a daytime sky. Then cirrostratus, the same ice spread into a veil over the whole sky, often so faint that a ring of light around the sun or moon is the only sign of it; the Met Office notes that the sun through cirrostratus still casts a shadow, which is how you tell it from what comes next. Altostratus follows, a grey sheet that lets the sun through as a watery disc and throws no shadow at all. Last comes nimbostratus, dark and featureless and thick enough to blot the sun out entirely. That is the one that rains. Cirrostratus is the stage worth timing, since the Met Office puts the reach of that veil at roughly the next twenty-four hours and reads a featureless one as a warm front likely to bring persistent rain inside a day.
Four minutes on telling one genus from another in a real sky. Watch for the shadow test that separates cirrostratus from altostratus, the hinge between rain tomorrow and rain in a few hours.
Two further readings cost nothing. A barometer falling steadily says the system is coming your way, and the needle agreeing with the sky is worth more than either alone. In 1857 the Dutch meteorologist C. H. D. Buys Ballot published the rule that turns wind into a map: stand with your back to it in the Northern Hemisphere and the low pressure lies to your left, which tells you where the centre sits and which part of the system is about to reach you. The rhyme about a red sky at night belongs in the same box, and the Met Office defends it on the merits, since British weather travels west to east, so a red sunset lights the back of a system already leaving and a red dawn the front of one arriving.
The honest limit is that all of this describes a front, and a front is not the only thing a sky can be doing. On a still summer day with no front for a thousand miles the clouds run the other way entirely: small fair-weather cumulus at mid-morning, growing taller through the afternoon as the ground heats, and by four o’clock a cumulonimbus with an anvil head and a squall underneath. You read that one by watching a single cloud grow instead of watching a procession pass. Learn both and you will be right more often than not, which is the most any forecaster claims. Howard did none of the physics, and his half is the half you actually use. Once you can say cirrostratus out loud, a white sky stops being weather in general and becomes a time of arrival.