D Kavita Singh majik schools, the Tamluk-Kalighat-Tribeni school which included Midnapur, Hooghly, Howrah, and 24 Pargana villages; the Birbhum-Kandi-Katwa school which included parts of Birbhum, south Murshidabad, and Burdwan; and the Murshidabad-Behrampur school. At the time of writing Ray did say that only the elders still remembered the samaj rules which were now being breached by the present genera- 93 tion. Ray's writings on this matter have come in for scathing criticism from the Visvabharati anthropologist, Benoy Bhattacharya, who has studied the patua community of Birbhum in some detail.6 According to Bhattacharya his informants had never heard of such a thing and he dismisses it as fiction. If Ray's talk of the samajik school is fiction it is a useful one, because the boundaries that he draws work very well with the stylistic idioms that one sees in museum collections. Even without a priori knowledge of Ray's writing, patas do begin to organize themselves in regional terms, identifiable as scrolls from Burdwan, Birbhum, Midnapur, Manbhum or Murshidabad. The following sections of this paper examine patas from three best-documented areas, areas that I call Midnapur, Birbhum and Murshidabad, but are coterminous with some cultural or kinship region and not the district borders themselves.7 Medinipur: (Illus. 8-11) The patuas of Medinipur are among the most active today; owing perhaps to their proximity to Calcutta, they receive the lion's share of the government and crafts bodies' patronage and many features of Medinipuri scrolls are mistaken as typical of all patas. The contemporary interests of the patuas, for instance, their readiness to deal with prime ministerial assassinations, the atomic bomb, the capsizing of a steamer, are typical of patuas from Medinipur rather than those of other regions. Oddly, their openness and inclusiveness on the level of subject matter, contrasts with a conservatism on the level of pictorial form. In its visual style, the Medinipur tradition is less eclectic, adhering to a traditional visual vocabulary which depends upon lakshanas (iconographic features) and mudras (stylized gestures), choosing not to incorporate contemporary imaging in their treatment of even contemporary themes. As physical entities, the scrolls from Medinipur are relatively narrow — usually less than two feet across, suited for easy handling by a single performer. The scrolls usually begin with a large panel of an icon or a figure in state, a socially prominent character, who is not necessarily the protagonist (Ramayana scrolls begin with an image of Dasharatha in court, Krishnalila scrolls start with an image of Kamsa). This is followed by many very narrow registers, occasionally interspersed with large dramatic panels. The registers are crowded with two or three events depicted in each, often obtruding into each other's space and across the border. In Krishnalila and Manasa Mangal scrolls this contravention of the border is transmuted into a marvellous device, where the river flows from one register to another, linking sequences of events that occur on its banks. The scrolls seem to end in medias res, with no formulaic panels recounting definitive conclusions.8 Numbers 27-28