24 BULLETIN 462, U. S. DEPARTMENT OF AGRICULTURE. 
to subirrigate lands of such type near Orlando, but in every case com- 
plete failure has followed. There have been also several attempts 
to subirrigate the heavy clay loams of Alabama, but with most dis- 
heartening results. The causes of failure are obvious in every case 
when the nature of subirrigation is considered. In the attempts 
made near Orlando the water was lost in the depths below the sur- 
face soil and would have had to fill up all the soil above the clay 
substratum before surface irrigation could take place. In the Ala- 
bama installations the clay surface soils were water-logged for a dis- 
tance of a foot or two on either side of the tile, the soil beyond being 
perfectly dry. 
In the Sanford district, however, the distribution of the water is 
not impeded by heavy soils. There the action of the water during 
subirrigation has been studied carefully with the aid of a soil- 
sampling apparatus. These tests show that the irrigation of the sur- 
face soils is not due so much to the action of capillarity as to the 
simple process of filling up the soils with water. Tests taken at the 
time of irrigation show that the soil reaches the saturation point 6 
inches under the ground surface before the surface soils are irrigated 
sufficiently. It also is common to find free water standing in the 
soils at 1-foot depths when the plants are in need of irrigation. 
These tests, along with many others, show conclusively the nature 
of subirrigation. 
It is evident that this system of irrigation requires large quanti- 
ties of water, and consequently subirrigation under soil conditions 
similar to those at Sanford is not an economical method where the 
water has to be pumped. Nevertheless, it is common for the uniniti- 
ated to regard this method with favor on account of its supposed 
saving of water. The large quantity of water needed is no particular 
drawback to the Sanford farmer because the nature of the. water 
supply assures him an unlimited quantity at no operating cost. 
A comparatively even ground surface, or a uniform slope, is re- 
quired for successful subirrigation. If the ground is uneven the low 
places will become flooded while the high knolls will not be irrigated 
sufficiently. 
Another factor in the successful operation of a subirrigation sys- 
tem is the correct kind of cropping. It has been stated that the San- 
ford district depends almost entirely on garden products. These 
crops invariably are shallow-rooted, most of the plant roots feeding 
within 6 inches of the surface. These, being annual plants, do not 
have an extensive root system. It has been demonstrated in the West 
that the subirrigation of trees is a failure on account of the stopping 
up of the tile by tree roots. This is especially true if the soil is rich 
in plant food to the depth of the tiling. On the other hand, there are 
records of successful subirrigation of citrus groves through tile sys- 
