66 SOCIAL SCIENTIST separately. Of the 170 odd pages in the book, more than 70 are devoted to tables. Perhaps the most valuable findings of the survey are those concerning the nature of the small peasant holding as a production unit. There is, firstly, the remarkable uniformity in cropping patterns across size-classes. The differences in cropping patterns (in particular the 'distress cropping pattern') observed in the Farm Management Surveys of the pre-^green revolution" days are no longer visible and "a'more or less uniform set of priorities is observed by farms of different sizes" in all the three^gricultural regions. Secondly, the traditional inverse relationship between size and productivity is seen to be considerably diluted and even reversed. For individual crop yields, the authors report "no evidence of a decline in crop yield with increase in farm size" for the state as a whole, although maize in Region III alone displays such a relation; on the other hand, wheat and rice in Region I are seen to be positively correlated with farm sizfe. For overall yield rates and productivity per acre the survey yields an interesting result, with Region I (the most advanced region) displaying increasing productivity with area operated, Region II showing the inverse relation and Region III showing no significant variation across size classes. Small farms continue to have an edge over the large ones in terms of cropping intensity, but per acre expenditure on inputs varies positively with size, an interesting feature being the very high share of expenditures on draught cattle for small farms. The irrigation base of small farmers also compares well with that of the large farmers. While this in itself is not a new featuie, what is more striking is the considerable amount of investment undertaken by small farmers. The authors find that "except in Region II, the per acre investment undertaken by the marginal and small farmers compares very favourably with the biggest farmers" and surmise that since this investment is taking place despite the deficits on current account it must be financed through borrowings. Thus, to sum up, the small peasant holding as a production unit has come to resemble the large holding fairly closely in terms of the crop mix, yields, irrigation etc, the major differences being in terms of expenditure on inputs and the ability to afford indivisibles (eg, tractors and other farm machinery). Coming to the patterns of income generation, the most significant finding concerns the importance of non-farm incomes. For the marginal farmers in particular, these incomes are overwhelmingly important, 60-70 percent of the total household income being derived from these sources. More generally, non-farm incomes are a major factor reducing the relative inequality of total household income across size classes of operated area. Thus, while the distribution of land is highly skewed, and the distribution of farm business incomes is even more skewed, income from non-farm sources is much more evenly distributed, so that the distribution of total household income is less skewed than it would have been with the farm business incomes alone. There is a strong