tic of showers and thunderstorms, while other forms of nimbus are more likely to be associated with cyclonic rain. Thus an entry of the form of cloud gives a moderately good idea of the character of the weather.

The example No. 9 of the plate is from London. The small dark raggy clouds, below a uniform cloak of lighter formless cloud are very characteristic of rain. It must be distinctly understood that when an observer is under cumulo-nimbus he only sees nimbus, and that he must judge whether to apply the epithet cumulo by the appearance of the rain cloud as it comes on or passes away.

10. Stratus (contracted Str.)—A thin uniform layer of cloud at a very low level. The example, No. 10 of the plate, is a typical specimen of the uniform type taken in the south of England, but the sky is often covered with detached masses of flat structureless cloud, which must also be reported as stratus.

A uniformly overcast sky, from which no rain falls, should be reported simply as overcast without applying any name. The word pallium, or cloak, which has sometimes been used, is objectionable, as the word suggests some definite structure of cloud. This it cannot do, for a uniformly overcast sky may either be verging on precipitation, and therefore almost nimbus, or else the under surface of stratus, which is usually associated with radiation and fine weather.

Fog is condensed vapour resting on flat ground; haze is thin fog. Sometimes fog may be watched rising off the ground and forming a stratum of well defined cloud very low down. This should then be reported as stratus.

When vapour condenses round a mountain top, obviously at some height above the valleys, the product is conventionally called cloud and not fog.

The above ten names-cirrus, cirro-stratus, cirro-cumulus, strato-cirrus, cumulo-cirrus, strato-cumulus, cumulus, cumulonimbus, nimbus, and stratus, are sufficient for ordinary purposes,

but the highest class of cloud observers may note some of the rarer and more transient forms as follows:

Cirro-filum* or cirrus stripes for that particular form of cirrus which takes the form of long lines instead of light irregular wisps.

Cirro-nebula or cirrus haze, for that high formless haze which gradually whitens the blue of the sky above, while the horizon remains clear. This sometimes developes without any cirrostratus, and produces halos round the sun or moon when not too dense.

Mammato-cumulus* or festooned cumulus, when the under surface of a cumulus droops downwards like a festoon of drapery. Figure 2 is an example of this form of cloud, from a sketch by Mr. Clouston.

Turretted cumulus, when the cumulus is found in a long line of little cloudlets like heads. This occurs usually before thunderstorms.

Cumulo-stratus, when the top of a cumulus becomes flattened out and spreads out wider than the base.

Mammato-stratus, when the under surface of a stratus cloud, is gathered up into drooping festoons.†

*** These terms are all due to Rev. C. Ley.

† For further details as to the nature of these clouds, see Abercromby, Weather," p. 116. London: K. Paul, Trench & Co., and Appleton & Co. New York.

It has already been mentioned that clouds-especially some forms of cirrus-form long lines in the sky, and that these stripes are often ribbed or striated. It has been discovered that very valuable deductions as to the future weather can sometimes be made by noting the direction in which the cirrus stripes lie. When the stripe is ribbed, the ribs may be either at right angles to the length of the stripe, or form a considerable angle with the lie of the stripe. The direction in which the ribs lie is unimportant. The direction of a cloud stripe is found as follows:

It is a law of spherical perspective,—that is of the perspective of curved lines such as cirrus stripes, which are parallel to the round surface of the earth—that parallel stripes, viewed end on,

H

N.W.

e

FIG. 3.-Cirrus stripes viewed end on. All converge in N., and therefore lie N. & S. The striæ of b are horizontal, and lie E. & W.; those of c converge in N.W., and lie from N. W. to S.E. a is the penniform type of cirrus; b & c are the ribbed or striated type; d&e are the wispy type.

H

appear as a series of slightly curved arches, converging to, or radiating from, a point a little below the horizon; and that the bearing of this point is the same as that of the direction of the stripes. For instance, a series of stripes, lying north and south, would appear, as in figure 3, like a series of slightly curved arches, converging to, or radiating from, a point a little below the north point of the horizon; with the exception of the stripe b, which passes directly overhead, and is therefore quite straight.

A similar, but not identical, series of stripes viewed crossways, or at right angles to their length, would appear like a series of low arches, whose summits bore in a direction, square, or at right angles, to the lie of the stripes. For instance, the perspective view of a series of stripes which lay north and south, viewed towards the west, would be as in figure 4, where the summits of

H

W

H

FIG. 4.-Cirrus stripes viewed broadside on.

The summit of the arches f, g, h all bear W., and therefore the stripes lie N. & S. The striæ of h converge on W., and lie E. & W.; those of y converge towards the N.W., and lie from N. W. to S. E.

all the arches bear due west. The general law is, that if any part of a cirrus stripe lies parallel to the horizon in one particular direction, the stripe lies square, or perpendicular to that direction that is to say, that if the stripe appears parallel to

the horizon, looking west, then the stripe is lying north and south.

Hence the direction in which stripes lie can be determined either by noting the bearing of the point from which they appear to radiate, or the bearing of the summits of the arches which they form.

The point from which cirrus appears to radiate is called the R point, or radiation point of the cirrus. The name is derived from the radiant point of meteors; or the point from which, as an effect of perspective, a stream of parallel meteors appear to diverge, or radiate from, as they shoot across the sky.

The direction in which the ribs or striæ lie is found on exactly the same principles. For instance, in fig. 3 the striations of b are parallel to the horizon looking north, and therefore they lie east and west; while the ribs of c converge on the north-west, and therefore they lie from N.W. to S.E.

In practice it is comparatively rarely that stripes can be seen of sufficient length to show much curvature. We see usually only a length, so short as to appear straight, but in most instances the curvature may be neglected altogether, and it is sufficient to note the point where the line of the cirrus produced mentally meets the horizon. This will be fully illustrated when we come to discuss the perspective of motion.

A very simple way of learning cloud perspective is to stand at the end of a long room, and to assume that you are looking due north. Then the lines of the two cornices on either side of the room will appear to converge towards the north, and if you suppose striated or ribbed lines, like those in fig. 3, to be painted on the ceiling lengthways to the room, they also will appear to converge in the manner shown in that diagram. Of course there is no trace of curvature in such short lines as these.*

* For explanation of the relation between the lie of a stripe and the centre of a cyclone, see Abercromby, "Weather," p. 92. London: K. Paul, Trench and Co.