82 DEVASTATION BY TORRENTS. 



■was first cleared, it required irrigation mucli less frequently than later 

 for annual crops, such as vegetables. Progressively with the years, 

 the land took in the irrigation water more slowly and the evaporation 

 from the soil was more and more rapid. The reason for this is that at 

 frst the soil was well filled with humus, which the sun and air grad- 

 ually burned out. In a small way and on parts of the ranch water run 

 in a very small stream on the ground will go furthest at first. The 

 soil deprived of humus seems to have to learn to take the water. Thus 

 you can take a little rill of water on hard ground for say fifty feet. 

 As the ground takes the water and softens it, the ground takes more 

 Vi-ater, until your rill will end at thirty-five feet. 



A fiomewhat contradictory case is that of the torrent bed of the 

 Whitewater on the desert. This stream flows out of the San Bernar- 

 dino mountains and takes its name from a sort of white clay it carries. 

 Out in its boulder bed it deposits this clay and makes a water-tight 

 channel, which extends the flow further and further out towards the 

 desert. This goes on until there is a heavy rain and flood flow. This 

 flow is short. It rolls the boulders about and breaks the clay bottom. 

 As soon as the flood flow is over, the permanent water flow is found to 

 be far back toward the mountain. This is the contrary of what we 

 would expect. It then goes back to zanja building and so again ex- 

 tends its flow desertward. 



In the usual experience of irrigation, the longer the time of flow, 

 the further the water goes to its limit. 



This ranch experience of the value of humus in getting land to 

 take and hold water, and its value in preventing on hill lands storm 

 off-flow is being availed of by bringing in and dressing the land with 

 humus-making manures or plowing in green crops. It is astonishing 

 to observe the difference that a good dressing of stable manure will 

 make in the water-taking capacity of clay land. 



OBSERVATIONS AT THE KINNELOA BANCH. 

 The Kinneloa ranch has its main orchards on mesas one to two 

 hundred feet above the general level on the three sides. The slope is 

 f"om one in eight and one-half feet to one in twenty-two. Before the 

 chaparral was cleared off, there was not a barranca, gully or wash on 

 the mesas, nor on the steep slopes from the mesas to the lower lands. 

 The question of the storm off-flow became important the second year 

 after the clearing. In spite of all the care taken to guard against soil 

 cutting by storm off-flow, several have occurred. One barranca and 



