Sleet consists of raindrops that are frozen while they fall. When the invisible moisture is frozen while it is condensing, instead of collecting in minute drops, it crystallizes and appears as snowflakes or snow. Snow crystals assume very beautiful forms. They consist of star-shaped collections of minute crystals. These crystals, however, only form in the absence of a strong wind.

Hail occurs when considerable differences of temperature exist between the higher and lower strata of air and the moisture is rapidly condensed. It is especially apt to occur in summer towards the close of an unusually warm day and during a thunderstorm. Hailstones consist of alternate layers of ice and snow, and are now generally believed to be caused by the whirling or rotation of the air around a horizontal axis, by means of which raindrops are carried through a cold snowcloud where they are both frozen and coated with a layer of snow. The successive coatings of ice and snow are due to the hailstones being carried through rain and snow clouds in turn, until they fall to the earth by reason of their increase in weight.

As a rule, hail falls during a violent thunderstorm. It falls in a narrow path or streak, while the rain covers a much wider area. This narrow streak marks the course of the vortex of the storm, within which the air is highly rarefied. Intense cold is thereby produced, and the moisture in the air is almost instantly frozen, while everything is whirling with great rapidity. Nearly every condition of the atmosphere may be passed through by the forming hailstones, whirling at one moment, among the snowflakes in the next moment and then again whirling in the frigid vortex.

Mr. E. Alexander Scott, of Philadelphia, gives the following description of one of the most remarkable hailstorms on record, which occurred in Philadelphia in the summer of 1870. This storm was noted, not only for its severity, but also for its long duration, and for the restricted area on which the hail fell.

It occurred on a Sunday at noon. The heavens darkened rapidly, as in the formation of a thunderstorm. The sky, however, presented a most unusual spectacle. Instead of the snowcloud rising from one quarter of the heavens, two portentous disturbances approached each other rapidly from opposite quarters, the one from the northeast and the other from the southwest. They advanced like great armies in battle array, huge masses of clouds tumbling over each other like the smoke of a battery of field guns.

Their advance lines were parallel and nearly straight. No rain was falling, and very little fell during the storm. When the contending hosts met, darkness, like night, overshadowed the city. The first scattering hailstones that fell struck the roofs of the houses with such a resounding thud that everybody ran to shut the window-blinds; but none was quick enough to prevent the destruction of glass that followed. For many weeks it was not possible to find enough glass in the country to repair the damage.

The stones were all large, a considerable proportion being the size of black walnuts. They lay piled up against the walls of the houses on the east and south sides of the streets, in some places eighteen inches deep.

The most remarkable feature of the storm was that the hail fell only on a streak about three-quarters of a mile wide, running from southwest to northeast in a straight line; no hail falling in other parts of the city, although it continued to fall for nearly twenty-five minutes. It required hours of hard work to clear the sidewalks.

One of the best evidences of its severity is shown by the fact that the hospitals were filled with persons who came, or were brought there, to have their wounds dressed; many with broken fingers, and multitudes with gashed faces and scalps. Horses generally were badly cut, and many runaways resulted; and in one case a runaway hearse tipped over and deposited its burden in the gutter.

This storm was accompanied by the electrical character

istics usual in thunderstorms, but they were not of an extraordinary character. The storm rather resembled a tornado, or whirling storm, the vortex of which, however, was not low enough to envelope the buildings in its destructive folds.

CHAPTER XVIII

OLD PROBABILITIES

"Old Probabilities" was the name given to the Chief of the United States Weather Bureau when the Department was first established by Congress. This name was originally given as a matter of pleasantry, for the predictions all began with: "It is probable;" and it was also believed that weather predictions, even when undertaken by the United States Government, could only be regarded as the more or less shrewd guesses made by the so-called weather prophets. General Myer, who organized the Weather Bureau, was much annoyed by the name, and changed the form to read: "The indications are," etc. The name, "Old Probabilities," though sometimes still given to the Chief of the Weather Bureau, is not employed as much now as it was formerly.

From the earliest times there has been a desire on the part of all kinds of people to know just what the weather will be during the next few hours, or the next day. This is natural, since our comfort is often greatly affected by marked changes in the weather. When it does not rain for long periods of time, the droughts may cause great loss to the wheat or other crops. On the other hand, if the rains are excessive, the lowlands of all large rivers may be visited by great floods, like those that, during the spring of 1897, overflowed the lower Mississippi Valley and, covering an area of some 13,000 square miles, caused a total loss in live stock and crops to a value exceeding $12,000,000. A violent storm at sea may cause even greater losses in vessels and lives. The prevalence of a hot, dry wind may blight the crops in

many entire States; while the unusual occurrence of frost in countries like Florida, may cause great money loss by completely ruining the orange crop.

It is impossible for man to prevent these sudden changes in the weather, but, were he able to foresee exactly when such changes were coming, he might take such steps as would either entirely avoid, or at least greatly lessen the damage. Efforts have been made to predict coming changes in the weather, by observing either the actions of certain animals, or certain appearances of the sun or moon, as well as in many other ways. Such predictions, however, are generally unreliable.

It certainly should be possible to predict coming weather changes. The movements of the wind, or of heated or cold air, the production of a storm, the falling of rain or snow, the coming of frost or of a hot wave, are all natural phenomena that are produced by definite causes. If one knew all of the causes and the exact order in which they acted, he ought to be able correctly to predict coming weather changes. Take the case of a rainstorm. If one knew that the air over a certain section of the country was hot and loaded with moisture, and was certain that changes were coming that would chill the air below its dew-point, he would be as safe in predicting a fall of rain as he would be the falling of an unsupported body to the earth. The difficulty is, however, that so many things might occur to prevent the air from being chilled; or, the moist air may be replaced by hot, dry air, so that, instead of the predicted rainstorm, there might be an unusually dry spell. Another difficulty arises by reason of changes occurring in some distant part of the earth, of which the weather prophet has no knowledge. If he could be in a thousand places at the same time, he might be better able to acquire reputation as a weather prophet.

The telegraph and telephone, and especially wireless telegraphy, enable the weather prophet to acquaint himself with actual weather conditions that exist at any time over