20 BULLETIN 502, U. S. DEPARTMENT OF AGRICULTURE. 
this point are available. Information of this character has been 
difficult to collect because it is almost impossible to measure the dis- 
charge of the well except by noting the increase of flow at the outlet 
of the tile as the wells are installed. This could be done readily if 
the wells were installed after the line was completed, but construc- 
tion in bad ground often necessitates the drilling of the wells as the 
work progresses, and this makes it impossible to determine how much 
water is developed by the tile line and how much by the wells. The 
following data were collected on a line of tile 350 feet long and 7 
feet deep: When completed, the drain without relief wells discharged 
3.2 gallons per minute. Six wells were installed in one day, imme- 
diately after which a measurement was made, which showed a dis- 
charge of 21.4 gallons per minute. Two of the six wells installed 
did not flow at all. Another line in this same system, and with the 
same length and depth, discharged 37.5 gallons per minute after 
completion. Twelve wells were installed in one day, after which the 
discharge was 85.7 gallons per minute. This latter example prob- 
ably is more nearly representative of average results. About 300 
feet of each of the above two lines were in shale. 
CONSTRUCTION. 
Construction of drainage systems in shale lands has varied greatly 
in respect to difficulty of installation and cost. . Much of the shale 
is quite hard, or contains hard concretions, which makes necessary 
the use of picks. Trenches where the greater depths are in solid 
shale stand well and need little or no bracing if the work is handled 
properly. Shale makes a very good foundation for laying tile, and 
the coarse, broken shale is good material for blinding the tiles 
Generally speaking, the work in shale is not difficult, but trenching 
in the saturated adobe soil is a real problem. Outlet lines usually 
have to pass for considerable distances through soil not immediately 
underlain by shale, and in many lines the upper several feet of the 
trench must be excavated through the soil before the shale is reached. 
With the exception of saturated fine sand.or quicksand, no class of 
material is more difficult to handle than adobe. When partly satu- 
rated it becomes sticky and adheres to the materials and tools with 
a tenacity that makes progress difficult and tedious. The skeleton 
spade is the only tool that will handle it with any degree of success. 
The ordinary shovel will not scour. When this soil becomes com- 
pletely saturated it often assumes a semifluid state that makes the use 
of tight sheeting necessary ; frequently, not only must the sides of the 
trench be sheeted, but also the face. 
The most successful cribbing in such material consists of two heavy 
timbers, held apart by trench jacks, behind which is driven the 
