-Ly 
COST OF CONSTRUCTION 3346 FEET OF V-BOTTOM 
CONCRETE BEDS —- CONSTRUCTED SPRING AND SUM 
MER 1945, 
Oontractor @ $1.50 per ft. of 
bench, material and labor $5,019.00 
Gravel " at $2. er yd. 275.00 
3" WER ae $2.05, bg per te 569" 68 
Copper expansion joints 20! apart 
at each 168,00 
Rental on forms at 20¢ per ft. 669, 20 
T! es rng charges on forns 40, 00 
Bulktead drain pipes and frame pipes 90,00 
§0, 528-88 
Cost per foot $1.95 
Above costs do not inolude cost of re=- 
placing heating coils, installing watering 
system, wiring frames or labor plaoing drain 
tile and gravel in beds, 
G.a\; ty 
SHE Hh Tt a H-1b Hb Hb tab HHO fe et 
JOHN R. THOMAS 
WHITFORD, PA. 
AUTOMATIC WATERING OF GROUND BEDS 
In developing a new technique we often 
advance as much through errors, learning 
what to avoid, as we do through our successes, 
60 I will therefore devote more time to ‘the 
mistakes we have made than I will to the ad- 
vantages of automatic watering with which we 
are all more or less familiar, 
Some five years ago we built two 120 ft. 
by 42 inch by 9 inches deep, ground beds 
made from concrete with a V~Bottom having a 
three-inch drop, This was our first mistake, 
The middle of the bed became saturated before 
the sides which caused the edges of the beds 
to run too dry. I know a shallower V would 
be better, and I believe a flat bottom with 
a water channel cast lengthways in the oon- 
orete would be better still. We made one bed 
level from end to end, but the other we made 
with a slope of 2 inches in 120 feet. This 
was our second mistake. The low end of the 
bed was always wetter than the high end. This 
because water would lie in the lower portion 
longer than in the upper part also there was 
2 inohes more soil in the lower end acting as 
a reservoir for reserve moisture. 
Our firet installation was made with 
three-inoh tile laid along the middle of the 
bed with a covering of 1/2 inch orushed stone 
laid level and deep enough to cover the tile 
and the bed then filled with soil. I am ine 
clined to believe this was a mistake. In ow 
experience the fine particles of soil filtered 
down into the orushed stone and before long 
had completely filled the epaces between the 
stones. We are sure that the water had more 
diffioulty in penetrating this aggregate 
laterally than through soil alone, 
Our next mistake was in being in too 
mich of & hurry when automatically watering 
these beds. The water was introduced through 
a stand=pipe at one emi of the 120 foot bed 
through a 3/4" hose at 50 lbs. pressure, A 
little common sense would have told us that 
this pressure in the tiles would break out 
first where the soil was loose and the conse= 
quence was very uneven moisture throughout 
the bed. We have found since that a very small 
stream of water entering the tile does a much 
petter job as it allows the soil to absorb the 
water rather than peat fe to force water into 
the soil, We now take to 5 hours to water 
a bed which we first tried to do in 30 min= 
utes. 
So much for our mistakes. I will now 
describe the measures we have taken to correct 
them. We first removed the top soil and dis= 
carded the mixture of stone and soil in the 
bottoms of the beds, we also discarded the 
tile. Across the V in the middle of the beds 
we laid sheets of linabestos or transite one 
foot wide and raised above the concrete by 
1/4" blocks of the same material so the sheets 
would not hug the concrete bottom too closely. 
We then filled the beds with soil. Half way 
along the bed with the 2" slope we built a 
dam 2" high in the bottom of the bed leaving 
a@ spill-way so that the water would flow into 
the lower half of the bed only after it had 
raised 2" 4n the upper half, NOTE, we still 
have to drain the lower half after each water- 
ing. 
We next installed a tank with a float 
valve in the end of each bed, similar to a 
toilet tank, with the float adjusted to hold 
the water level 6" below the surface of the 
soil. Water is piped to the tanks so to 
water a bed all that we do is turn on a valve 
ani go about some other job. We leave the 
water turned on until the tensiometers begin 
to drop which takes from 4 to 5 hours. In 
another 8 to 10 hours the tensiometers have 
dropped to about 1 inch of vacuum which is 
the lowest point they reach, The high point, 
when we feel the beds should be watered, 
varies with the crop. We water at 5 inches 
of tension for Snaps and about 3 inches for 
Pompons and Tomatoes. We are old fashioned 
@nough to believe that a fluctuation in the 
moisture content of the soil is beneficial 
to root growth. 
We find that with repeated automatic 
watering the soluble salts concentrate in the 
upper 3" of soil, so once a month we give the 
beds a good surface watering. This often 
coincides with our feeding program. We find 
that the Nitrates are almost oompletely lost 
in the lower soil, Potash not quite so much 
while the Phosphates stay in place, 
Conclusions:- Beds must be level from 
end to end. Beds should have a very shallow 
Vor a flat bottom. If the bottom 1s flat a 
channel should be cast in the conorete s0 
that excess water oan drain away from the soil, 
We have never tried a layer of sani in 
the bottom of the bed to conduct water later 
ally but believe it would be better than 
gravel or crushed stone, We are now getting 
Con't on pg. 16 
