- Ii 
season and the amount of bean seed the farmer had left to plant. 3/ The system ofthe subsistence farmers in this area includes maize, beans and sorghum planted together at the same ti4e-in a number of similar arrays. Thr(.ugh the use of twin or double rows of maize and sorghum 4/ and a reduction in the n6pulaton of-beans 7hich con'sune the majority of planting time, productivity of planting labor and of bean seed was raised significantly by allowing each farmer to plant more land than he previously had been able to with his traditional cropping system. This non.-traditional technology is possible because amount of land is not a limiting fact r for most farmers in the area.

     Results from the Farmers' Tests in 1976 indicate that- on the average, each farmer could plInt about 40% more land using the same amount of planting labor and somewhat less bean seed and produce 75% more maize,
410% more sorghum>o the :snme. .amount of beans and 33% more income (Hildebrand and Cardona, 1977).The system allows him to work about 60 more days on his farm than other-wise would be the case and earn about.$.25 per day which is slightly under what he has to pay for hired labor. The productivity of labor for planting and bean seed (the scarcest resources) rose from $5.42 per dollar invested to $8.73, an increase of nearly 60%. Risk of loss is very low and there is no requirement for pesticides or fertilizer that the farmer normally does not us, in these conditions.

     In the Central highlands, another survey showed that land was the most limiting factor and capital was very scarce, but labor was relatively abundant throughout th. year. In addition, three strata of subsistence farmers were defined (Duarte, et.al. 1977). One groupp cannot produce enough maize to sustain the family, for the year, a second stratum achieves self sufficiency at times but not always, and the third always produces enough to satisfy family needs. Each of these three grous has different requirements even though their croppin- systcis basically the same, and a special technology was designed for each.
3/ In this case, the most frequent response frmcrs mad:
   with respect to bean yields was so-many pounds par
   pound of seed used,
4/ Details on the use of double rows cn bc. found in
   Hildebrand, 1976' Hildebrand, ct.al.  1977   iildchr-and
   and Cardona  1977, and French and Ilildcbrand 197C.