3. The utility of =Melted Snow and Ice is= obviously very limited. Moreover, its use is not free from danger if the ice is derived from contaminated water. Outbreaks of enteric fever have been traced in the United States to the taking of ice obtained from impure water.
4. =Rain-water= is a much more important source of water supply, and after passing through the soil it constitutes the chief part of the water we drink. The term, however, is properly restricted to the water collected immediately after its descent from roofs, etc. Its purity depends on three conditions—the character of the air it passes through, the cleanliness of and absence of lead from the channels through which it runs, and the condition of the water-butts in which it is stored. Rain-water is soft; in fact, too soft to be pleasant to the palate. In passing through the air, it carries with it a certain proportion of its constituents; in towns especially ammonia, soot, etc.; near the sea, it generally contains some salt; and being soft and having dissolved oxygen from the air, it dissolves an appreciable amount of lead from roofs or gutters.
The Rivers Pollution Commissioners found that out of eight samples of stored rain-water only one was fit to drink. They came to the conclusion that rain-water, collected from the roofs of houses and stored in underground tanks, is “often polluted to a dangerous extent by excrementitious matters, and is rarely of sufficiently good quality to be used for domestic purposes with safety.” Also, that in Great Britain, and more particularly in England, we shall “look in vain to the atmosphere for a supply of water pure enough for dietetic purposes.”
The use of rain-water for drinking purposes is only justified in isolated country houses where no better source is available; and under these circumstances the greatest care should be taken to prevent contamination with lead or organic impurities.
The amount of water falling on any impervious material obtainable from rain can easily be estimated, if the amount of rainfall and the area of the receiving surface are known. The average annual rainfall in this country is 33 inches (see page 236).
We may assume the amount practically available to be 20 inches per annum, and the area of the receiving surface 500 square feet. Multiply the area by 144, to bring it into square inches, and this by the rainfall, and the product gives the number of cubic inches of rain which fall on the receiving area in a year. One cubic foot, or 1,728 cubic inches, of water being equivalent to 6·23 gallons, the number of gallons of water can be easily calculated. To calculate the receiving surface of the roof of a house, do not take into account the slope of the roof, but merely ascertain the area of the flat space actually covered by the roof. This may be done roughly by calculating the area of all the rooms on the ground floor, and allowing an additional amount for the space occupied by the walls. It has been estimated that, even if a rain-water supply for towns were desirable, the amount collected from the roofs of houses would scarcely average two gallons per person daily—assuming the average rainfall to be 20 inches, and that there was a roof area of 60 square feet for each individual.