Though Russell has been researching rape for over ten years--the last five on behalf of the National Institute of Mental Health--her lengthy report on rape-in-marriage yields disappointingly few definitive results. To begin with, her door-to-door ""random sample"" of 930 San Francisco women (with high refusal and not-at-home rates) hardly seems typical of the country at large, and perhaps not even of urban centers--given (as she acknowledges) San Francisco's unusual population. On many issues, moreover, Russell either has to validate her guesses from the work of previous researchers (whose limitations she is quick to air), or explain that percentages only reflect information volunteered by interviewees--and might have registered differently if all respondents had been questioned on the issue. (E.g., almost one-quarter of the ""husband-rapists"" appeared from their wives' descriptions to be habitual drinkers--but since the respondents were never actually questioned on the subject, the data ""can be assumed to greatly underestimate the consumption of alcohol and other drugs."") The conclusions, then, must be regarded with some caution. Russell, like most feminists, traces wife rape to the roots of our patriarchal society and the antiquated concepts of wife as property. One in seven married women, she maintains, has been raped by her husband ""at least once""; and she is impatient with counterclaims of husband rape, or the attempt to see wife rape within the broader context of family dysfunction, because ""violence involving adults is primarily a male problem."" Following the example of the law, she defines three kinds of rape: by force, by threat of force, or by taking advantage of a helpless victim (drugged, asleep, etc.). It's not always clear, however, whether women who surrender because they fear reprisal have been given sufficient cause to feel threatened. The relationship between rape and beatings is also problematical: 49 percent of all abused wives are beaten only, 14 percent are raped only, and 36 percent experience elements of both--so the most Russell can conclude is that the rape-only victims frequently get lost in the shuffle. She is able to isolate almost no outstanding social characteristics of the husband-rapist (he appears to cut across classes, races, etc., rather neatly), or of the wife-victim (non-traditionalists predominate slightly, but it's the traditionalists who stick with the marriage). Within the patriarchal context, rape by an ex-husband is seen as a ludicrous attempt at ""repossession."" The police, of course, are characterized as largely biased on behalf of the husbands. In addition to quoting from 89 rape victims, Russell provides some legal case rundowns and lists of states with favorable and unfavorable laws. Ultimately, though, the study raises more issues than it has the expertise to handle.