58 SOCIAL SCIENTIST colonial Raj the 'surplus' was appropriated from the agrarian classes by the state in the name of 'revenue'. Undoubtedly, a specific form of ownership and control of the means of production is always in correspondence with a form of appropriation of surplus labour and, therefore, it is logical to take at first the pattern of landownership in Sultanpur district of east U P. As Table I shows, during 1945-1977 the percentage of ownership of landholdings has decreased in all size-groups, expect the first group (upto one acre) which shows a sharp increase from 39.90 to 60.79. Similarly, there occurred structural changes in the area held by various size-groups. The area under the first group more than doubled, from 7.77 per cent in 1945 to 17.28 per cent in 1977. The area under the second group remained constant while that under the third group increased from 18.18 per cent to 24.38 per cent. On the contrary, the area under the fourth and fifth groups decreased from 19.59 per cent to 13.29 per cent and from 9.04 percent to 7.90 per cent, respectively. The area under the sixth group was almost constant (from 5.02 per cent to 5.38 per cent), while that under the seventh and eighth groups of large landholders drastically decreased from 14.05 per cent to 9.91 per cent and from 7.71 per cent to 3.23 per cent, respectively. TA iBLE ] [ PATTER] N OF LAND( 3WNERSHIP IN SULTA NPUR DISTRI CT (1945-19: 77 Size Groups Percentage of Landholdings % area held Average area per (acres) holding [acres) 1975 1977 1945 1977 1945 1977 Upto 1 39.90 60.79 7.77 17 28 0.58 0.50 1-3 29.60 18-63 1864 1863 1.87 1.76 f 3-5 13.82 12.79 18.18 24.38 3.90 3.34 5-8 9.23 4.01 19.59 13.29 6.29 5,82 8-10 3.01 1.67 9.04 7.90 890 8.31 10-12 1.35 0.88 5.02 5.38 10.96 10.71 12-25 2.59 1.08 14.05 9.91 16.10 16 17 Above 25 0.50 0.17 7.71 3.23 45.46 33.80 SOURCE: (a) U P Zamindari Abolition Committee Report (1948) pp 34-39, (b) District Economics and Statistics Office, Sultanpur, Bulletin (1981) p 9, (unpublished date). NOTE: Due to the nature of the data avilable for two periods, the break-up of size-groups is not uniform. Thus we can say that ihe pattern of landownership is changing. However, the situation at this stage is quite puzzling. Nevertheless, the changing pattern becomes more manifest if we reduce all the above eight size-groups into three classes: (i) upto 5 acres, (ii) 5 to 10 acres, and (in) above 10 acres. Then we notice the following 'trend* of land-ownership in this district: (a) in the first class the percentage of holdings increased from about 83 to 92 and their area, too, increased from slightly more than 44 per cent to 60 per cent; (b) the percentage