causes the condensation and subsequent reappearance of the evaporated water in the form of rain or snow.

We now have all the conditions necessary to the creation of a flowing stream of water. As soon as the ability of the soil to absorb the rain or melted snow is exhausted, it begins to flow along the inclined surface, always seeking its lowest level, until the union of one small stream with another in a common channel produces a creek or river of importance.

ABOUT RAINFALL IN GENERAL.

Owing to the irregularity and unevenness of the surface of the earth, causing

differences in the area and elevations of watersheds, and variations in the temperature of the atmosphere together with the varying distances between bodies of water, the length, size and the volume of flow of streams at different seasons of the year is a matter of much uncertainty. It depends upon the amount of rainfall on the area tributary to the stream and its distribution throughout the year either by the constant recurrence of rains or the melting of accumulated snows. In the case of large rivers, where the volume of flow is derived

from a large watershed having a copious rainfall, the supply is probably so much greater than all possible demands, that investigations of the rainfall are unnecessary But such rivers do not frequently exist in arid regions, and the irrigation engineer is generally called upon to obtain a supply from streams having a limited as well as a poorly and unevenly watered drainage basin.

In such cases it is necessary to proceed with the utmost caution, and to carefully investigate the amount and distribution of

the rainfall throughout a sufficient number of years to determine the available supply of the stream.

VALUE OF STATISTICS.

All civilized countries have statistics of the rainfall on its principal river basins and watersheds and of the flow of its principal streams at different seasons of the year. The value of these statistics, of course, depends largely upon the efficiency of the service under which they are prepared, but it is safe to recommend the reports of the signal service and meteorological departments and the reports of the geological survey of most countries, as a

source from which a large amount of reliable information can be drawn regarding the capacity and discharge of the principal streams of those countries.

Statistics prepared by private individuals and corporations, who are or have been the projectors and owners of water works or irrigation enterprises, may be said to be equally as valuable if not more so than those prepared under government supervision, and when they are obtainable much valuable information can be obtained from them. But as a rule these only relate to watersheds or streams which have

already been improved or appropriated, and are useful only in making examinations and reports on existing works, or for drawing conclusions and making estimates ity. No tables of statistics relating to on works projected in the immediate vicinrainfall or the areas of drainage basins of streams will be reproduced in this work, as this would simply be copying public records, which are open to all, and would

Occupy space which is more valuable for the discussion of principles.

(To be continued.)

NOTE

[OTE-[As the publication of this series has been suspended some fifteen months the reader may have to go back over some of the ground passed over before fully understanding this. The first seven chapters were mainly introductory, dealing with the errors of the early irrigators to a large extent because experience has shown that almost every one if left to himself will follow exactly in their footsteps. The study of error is as valuable as the study of going right, and every one should read the history I there gave. Those chapters contained also a large amount of facts necessary to a full comprehension of what is to follow and too numerous to repeat. Those who have not read them must therefore attend more closely to what follows, for it will be the more practical part, or how to do it as distinguished from how you don't do it.]

Next to the quantity of water at your service, the size of the irrigating head in which you can have it, and the length of time you can allow it to run, the slope of the ground becomes the most important factor in determining the method of applying water. If the slope is great you can not flood by checks of any reasonable size. If they are very small their number becomes a nuisance in making and in hand. ling the water from one to the other. If not small then the water stands too deep and too long in the lower part; and too much puddling, with compression of the soil, is the consequence, besides uneven wetting. All these are to be carefully avoided where possible. The slope may be so great as to compel you to terrace if the nature of the product will justify the expense. If the soil is easily worked it may pay to terrace some on a very light slope, making the terraces very broad. may pay even for alfalfa. If terracing will not pay you may then be driven to the use of basins or rings around the tree or vine where the slope is great. On such ground you can do little with large heads of water for a short run but must in some way work with small streams with a long

This

if

Good drainage in irrigation is almost as essential as water. Where the drainage is bad you must avoid flooding if possible and use small streams or you may have a sour, cold soil, with alkali perhaps appearing on the surface when it dries and becoming constantly worse. On the other hand, if the ground contains alkali, which rises to the surface in quantity sufficient to be injurious, flooding is often the only way it can be removed. If there is sufficient slope to carry water rapidly off, ground that has become badly alkalied may be put in good condition by a quick flooding that will dis. solve all the alkali on the surface and then letting it run off as fast as possible. And very liable to alkali it should be laid out so that it can be flooded in this way if necessary. But drainage must still be provided if possible. You must remember always that in irrigation, however slowly or carefully done, if done on a scale large enough for a commercial success, a much greater quantity of water is put upon the ground in a given time than is usually done by the clouds even when unusually generous. And in the course of the year one who has any kind of a respectable water supply will generally put more into the ground and have more run off of it than will be the case from rain in the wettest parts of the country. Therefore there must be provision for the water to run away beneath if possible, or, if the soil is not naturally well drained, care must be used in putting water into it so that there be no excess. Waste water ditches above ground must be provided for what runs away and all rights of way should include these for individuals as well as companies.

There is a distinction between "alkali land" and land liable to alkali that should be kept in mind, though the difference is

none too easy to define clearly. Almost any land forms some alkali in every country, and under bad drainage may show too much when irrigated. The spots of natural alkali seen on the surface of the ground in some of the deserts are instances of this. They are not found where the soil is perfectly open below with good water beneath. There is somewhere a subsoil of some tight material that stops the alkali as it is carried down by the water from some occasional cloudburst or unusual rain. Then the evaporation brings it again to the surface where it is left as a white powder. Alkali is not often in the

Oranges do not seem injured by considerable of it, though ground so badly drained as to make alkali possible' will not generally produce a fine orange. But the possibility of alkali is always to be considered in irrigation; for under ordinary treatment it grows worse instead of better. Thousands of acres in California that were once fertile land, showing no trace of alkali on the surface, but having it in the subsoil, have had it brought to the surface by bad irrigation and defective cultivation and drainage and are now about worthless though they may be reclaimed. But ground with an open soil and good sheet

SPECIMEN OF VERY GOOD FURROW IRRIGATION.

water beneath, and hardpan lands with a slope of fifty feet or more to the mile have been worked for years with the worst kind of irrigation without showing a trace of alkali.

Your choice of methods will also depend upon your object in irrigating. Almost every rule and caution that this work contains may in some places and for some purposes be disregarded. Are you irrigating for profit or only to raise something for your own use? If for profit it may pay you to do the best work possible. For remember that good irrigation is often as far superior in results to bad irrigation as