13. Heavy Rains. When very heavy rains occur it is desirable to measure immediately on their termination; and it will be found a safe plan, after measuring, to return the water to the gauge, so that the morning registration will not be interfered with. Of course, if there is the slightest doubt as to the gauge holding all the falls it must be emptied, the amount being previously written down.
14. Snow. In snow three methods may be adopted; it is well to try them all:—(1) Melt what is caught in the funnel by adding to the snow a previously ascertained quantity of warm water, and then deducting this quantity from the total measurement, enter the residue as rain. (2) Select a place where the snow has not drifted, invert the funnel, and, turning it round, lift and melt what is enclosed. (3) Measure with a rule the average depth of snow, and take one twelfth as the equivalent of water. Some observers use in snowy weather a cylinder of the same diameter as the rain-gauge, and of considerable depth. If the wind is at all rough all the snow is blown out of a flat-funnelled rain-gauge.
15. Overflow. It would seem needless to caution observers on this head, but as a recent foreign table contains six instances in one day, in which gauges were allowed to run over, it is evidently necessary that British observers should be on the alert. It is not desirable to purchase any new gauge of which the capacity is less than four inches.
16. Second Gauges. It is often desirable that observers should have two gauges, and that one of them should be capable of holding eight inches of rain. One of the gauges should be registered daily, the other weekly or monthly, as preferred, but always on the 1st of each month. By this means a thorough check is kept on accidental errors in the entries, which is not the case if both are read daily.
17. Dew and Fog. Small amounts of water are at times deposited in rain-gauges by fog and dew. They should be added to the amount of rainfall, because (1) “they tend to water the earth and nourish the streams,” and not for that reason only, but (2) because in many cases the rain-gauges can only be visited monthly, and it would then obviously be impossible to separate the yield of snow, rain, &c.; therefore, for the sake of uniformity, all must be taken together.
18. Doubtful Entries. Whenever there is the least doubt respecting the accuracy of any observation, the entry should be marked with a ?, and the reason stated for its being placed there.
Obs. The height at which the rain-gauge is elevated from the ground is a matter of considerable moment. Thus, one observer found the fall of rain at York for twelve months (1833-1834) to be—at a height of 213 feet from the ground, 14·96 inches; at 44 feet, 19·85 inches; and on the ground, 25·71 inches.
Later experimentalists have confirmed this curious fact. Thus, Colonel Warde found the following to be the relative rainfall at different periods for the four years extending from 1864 to 1867:
| Inches. | |||
| On a level with the ground | 1·07 | ||
| At a height of | 2 | inches | 1·05 |
| ” | 6 | ” | 1·01 |
| ” | 1 | foot | 1·00 |
| ” | 2 | feet | 0·99 |
| ” | 3 | ” | 0·98 |
| ” | 5 | ” | 0·96 |
| ” | 10 | ” | 0·95 |
| ” | 20 | ” | 0·94 |
One of the causes that have been assigned for this singular phenomenon has been—the greater exposure in elevated situations of the rain to dispersive action of the wind, a surmise which derives some support from the circumstance, that when a rain-gauge is placed on a building, the roof of which is flat, of large area, and with few, if any, chimneys to disturb the air currents, an amount of rain is collected equalling that obtained on the surface of the ground.