"There were some honest mistakes, and then there were some mischievious mistakes which happened because people wanted them to happen," Graham said Tuesday night.

During the election, Thompson campaign manager Francis R. Stone claimed that misplaced stickers were the work of Graham supporters, who he says used tape, black ink--and stickers--to deface Thompson's name in the ballot book, making it appear as if there were no Democratic nominee.

In addition to improprieties at the polls, Graham's campaign officials also charge that votes were inaccurately counted. They say election officials in some precincts removed stickers incorrectly placed on the ballots themselves, counting them as blanks or as votes for Thompson.

"This is one of the disadvantages of stickers," said Kimberling. "That's why many states don't allow them, because people don't know what the hell they're doing."

If Graham takes her complaints about the voting procedures to court, she may have a hard time proving her case. Although Thompson received little more than 50 percent of all ballots cast--leaving open the possibility that nearly half of the voters intended to vote for Graham--in past elections it has been common for voters not to express a preference in local races.

The best case scenario for Graham would be that the 2870 blank ballots--more than 17 percent of the total ballots cast--at some point were marked with a preference for her. Add to this total the number of Thompson votes which Graham claims might have been meant for her, and the outcome could go either way.

The number of blank ballots this year, however, is well within the range of past Cambridge state representative elections. When Graham ran unopposed in 1986, for example, more than 32 percent of the voters left their ballots blank. Even in contested elections, it is not uncommon for 10 to 15 percent of the ballots to be left blank.